Computational Models of Addiction: Formalization, Validation, and Limits
This section is represented as a source-backed SciDEX wiki artifact. The full source remains in the original computational-review repository.
Source: https://github.com/AllenNeuralDynamics/ComputationalReviewAddiction/blob/43d33061a22aa4fed3f7dc71fabbfa48ab1d6f48/content/09_computational_models.md
Citation anchors captured: 133
Citation contexts
-
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference The systematic translation failures documented in{ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the... -
2CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference The systematic translation failures documented in{ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the... -
3CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference The systematic translation failures documented in{ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference The landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest... -
2CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference The landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest... -
4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference The landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field 1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference The landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference0. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest... -
2CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference1 The landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference2. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest... -
2CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference3 The landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference4. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest... -
2CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference5 The landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference6. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest... -
2CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference7 The landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference8. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest... -
2CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference9 The application of RL theory to addiction rests on the foundational observation that midbrain dopamine neurons encode reward prediction errors (RPEs) consistent with TD-learning algorithms 3CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference0. This mapping between a computational variable (the RPE) and a neural signal (phasic dopamine) provided a formal vocabulary for describing how drugs of abuse could hijack learning mechanisms: if addictive substances... -
3CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference1 The application of RL theory to addiction rests on the foundational observation that midbrain dopamine neurons encode reward prediction errors (RPEs) consistent with TD-learning algorithms 3CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference2. This mapping between a computational variable (the RPE) and a neural signal (phasic dopamine) provided a formal vocabulary for describing how drugs of abuse could hijack learning mechanisms: if addictive substances... -
3CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference3 The application of RL theory to addiction rests on the foundational observation that midbrain dopamine neurons encode reward prediction errors (RPEs) consistent with TD-learning algorithms 3CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference4. This mapping between a computational variable (the RPE) and a neural signal (phasic dopamine) provided a formal vocabulary for describing how drugs of abuse could hijack learning mechanisms: if addictive substances... -
3CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference5 The application of RL theory to addiction rests on the foundational observation that midbrain dopamine neurons encode reward prediction errors (RPEs) consistent with TD-learning algorithms 3CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference6. This mapping between a computational variable (the RPE) and a neural signal (phasic dopamine) provided a formal vocabulary for describing how drugs of abuse could hijack learning mechanisms: if addictive substances... -
3CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference7 The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits 3CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference8. Lesaint et al. (2014) showed that a combination of model-based and revised model... -
3CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference9 The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference0. Lesaint et al. (2014) showed that a combination of model-based and revised model... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference1 The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference2. Lesaint et al. (2014) showed that a combination of model-based and revised model... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference3 The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference4. Lesaint et al. (2014) showed that a combination of model-based and revised model... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference5 The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference6. Lesaint et al. (2014) showed that a combination of model-based and revised model... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference7 The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference8. Lesaint et al. (2014) showed that a combination of model-based and revised model... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference9 The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference0. Lesaint et al. (2014) showed that a combination of model-based and revised model... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference1 The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference2. Lesaint et al. (2014) showed that a combination of model-based and revised model... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference3 The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference4. Lesaint et al. (2014) showed that a combination of model-based and revised model... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference5 Yet the MF/MB framework faces a fundamental challenge. Morris and Cushman (2019) argued that much compulsive drug-seeking behavior is not truly habitual but rather represents goal-directed pursuit of drug rewards — behavior that appears inflexible because the goal itself (obtaining the drug) dominates decision-making, not because the underlying computational architecture has shifted 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference6. This is not merely... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference7 Yet the MF/MB framework faces a fundamental challenge. Morris and Cushman (2019) argued that much compulsive drug-seeking behavior is not truly habitual but rather represents goal-directed pursuit of drug rewards — behavior that appears inflexible because the goal itself (obtaining the drug) dominates decision-making, not because the underlying computational architecture has shifted 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference8. This is not merely... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference9 Yet the MF/MB framework faces a fundamental challenge. Morris and Cushman (2019) argued that much compulsive drug-seeking behavior is not truly habitual but rather represents goal-directed pursuit of drug rewards — behavior that appears inflexible because the goal itself (obtaining the drug) dominates decision-making, not because the underlying computational architecture has shifted 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference0. This is not merely... -
2CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference1 The dominant view frames addiction as a shift from model-based to model-free (habitual) control 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference2. However, Morris and Cushman (2019) challenge this directly, arguing that compulsive drug seeking can be goal-directed yet appear inflexible 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference3. The resolution may depend on distinguishing different stages of addiction: early drug use may involve MB processes that become incr... -
2CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference4 The dominant view frames addiction as a shift from model-based to model-free (habitual) control 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference5. However, Morris and Cushman (2019) challenge this directly, arguing that compulsive drug seeking can be goal-directed yet appear inflexible 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference6. The resolution may depend on distinguishing different stages of addiction: early drug use may involve MB processes that become incr... -
2CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference7 The dominant view frames addiction as a shift from model-based to model-free (habitual) control 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference8. However, Morris and Cushman (2019) challenge this directly, arguing that compulsive drug seeking can be goal-directed yet appear inflexible 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference9. The resolution may depend on distinguishing different stages of addiction: early drug use may involve MB processes that become incr... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference0 The dominant view frames addiction as a shift from model-based to model-free (habitual) control 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference1. However, Morris and Cushman (2019) challenge this directly, arguing that compulsive drug seeking can be goal-directed yet appear inflexible 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference2. The resolution may depend on distinguishing different stages of addiction: early drug use may involve MB processes that become incr... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference3 Several recent developments have enriched the RL landscape beyond the simple MF/MB dichotomy. Russek et al. (2016) introduced the successor representation (SR) as a bridge between model-free and model-based systems, demonstrating that the SR — a predictive state representation combined with TD learning — can produce a subset of behaviors associated with model-based learning while requiring less computation than full... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference4 Several recent developments have enriched the RL landscape beyond the simple MF/MB dichotomy. Russek et al. (2016) introduced the successor representation (SR) as a bridge between model-free and model-based systems, demonstrating that the SR — a predictive state representation combined with TD learning — can produce a subset of behaviors associated with model-based learning while requiring less computation than full... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference5 Several recent developments have enriched the RL landscape beyond the simple MF/MB dichotomy. Russek et al. (2016) introduced the successor representation (SR) as a bridge between model-free and model-based systems, demonstrating that the SR — a predictive state representation combined with TD learning — can produce a subset of behaviors associated with model-based learning while requiring less computation than full... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference6 Akam and Walton (2020) reported findings unexpected under the standard RPE account: dopamine signals in certain contexts appear to carry information beyond simple model-free prediction errors, suggesting dopamine’s involvement in model-based RL through mechanisms where dopamine neurons encode either state prediction errors or a combination of reward and state prediction information 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference7. Shouval et a... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference8 Akam and Walton (2020) reported findings unexpected under the standard RPE account: dopamine signals in certain contexts appear to carry information beyond simple model-free prediction errors, suggesting dopamine’s involvement in model-based RL through mechanisms where dopamine neurons encode either state prediction errors or a combination of reward and state prediction information 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference9. Shouval et a... -
4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference0 Akam and Walton (2020) reported findings unexpected under the standard RPE account: dopamine signals in certain contexts appear to carry information beyond simple model-free prediction errors, suggesting dopamine’s involvement in model-based RL through mechanisms where dopamine neurons encode either state prediction errors or a combination of reward and state prediction information 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference1. Shouval et a...
-
4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference2 Akam and Walton (2020) reported findings unexpected under the standard RPE account: dopamine signals in certain contexts appear to carry information beyond simple model-free prediction errors, suggesting dopamine’s involvement in model-based RL through mechanisms where dopamine neurons encode either state prediction errors or a combination of reward and state prediction information 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference3. Shouval et a...
-
4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference4 The application of RL models to clinical populations beyond substance use disorders has produced a range of findings. Cockburn and Holroyd (2010) explored how asymmetric dopamine-mediated learning could account for features of ADHD 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference5. Li et al. (2014) found that individuals with psychosis updated reward values more rapidly, consistent with elevated prediction error signaling 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference6. Balasubramani et a...
-
4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference7 The application of RL models to clinical populations beyond substance use disorders has produced a range of findings. Cockburn and Holroyd (2010) explored how asymmetric dopamine-mediated learning could account for features of ADHD 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference8. Li et al. (2014) found that individuals with psychosis updated reward values more rapidly, consistent with elevated prediction error signaling 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference9. Balasubramani et a...
-
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference0 The application of RL models to clinical populations beyond substance use disorders has produced a range of findings. Cockburn and Holroyd (2010) explored how asymmetric dopamine-mediated learning could account for features of ADHD 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference1. Li et al. (2014) found that individuals with psychosis updated reward values more rapidly, consistent with elevated prediction error signaling 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference2. Balasubramani et a... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference3 The application of RL models to clinical populations beyond substance use disorders has produced a range of findings. Cockburn and Holroyd (2010) explored how asymmetric dopamine-mediated learning could account for features of ADHD 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference4. Li et al. (2014) found that individuals with psychosis updated reward values more rapidly, consistent with elevated prediction error signaling 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference5. Balasubramani et a... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference6 The application of RL models to clinical populations beyond substance use disorders has produced a range of findings. Cockburn and Holroyd (2010) explored how asymmetric dopamine-mediated learning could account for features of ADHD 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference7. Li et al. (2014) found that individuals with psychosis updated reward values more rapidly, consistent with elevated prediction error signaling 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference8. Balasubramani et a... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 2CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference, 4CitationThe landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...content/09_computational_models.md:line 6Open reference9 The application of RL models to clinical populations beyond substance use disorders has produced a range of findings. Cockburn and Holroyd (2010) explored how asymmetric dopamine-mediated learning could account for features of ADHD 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference0. Li et al. (2014) found that individuals with psychosis updated reward values more rapidly, consistent with elevated prediction error signaling 1CitationThe systematic translation failures documented in {ref}sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference1. Balasubramani et a... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference2 A fundamentally different class of models formalizes the transition from positive to negative reinforcement that characterizes the progression from recreational drug use to dependence. The opponent process framework, most extensively developed by Koob and colleagues, posits that chronic drug exposure recruits brain stress systems — particularly corticotropin-releasing factor (CRF) signaling in the extended amygdala... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference3 A fundamentally different class of models formalizes the transition from positive to negative reinforcement that characterizes the progression from recreational drug use to dependence. The opponent process framework, most extensively developed by Koob and colleagues, posits that chronic drug exposure recruits brain stress systems — particularly corticotropin-releasing factor (CRF) signaling in the extended amygdala... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference4 A fundamentally different class of models formalizes the transition from positive to negative reinforcement that characterizes the progression from recreational drug use to dependence. The opponent process framework, most extensively developed by Koob and colleagues, posits that chronic drug exposure recruits brain stress systems — particularly corticotropin-releasing factor (CRF) signaling in the extended amygdala... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference5 A fundamentally different class of models formalizes the transition from positive to negative reinforcement that characterizes the progression from recreational drug use to dependence. The opponent process framework, most extensively developed by Koob and colleagues, posits that chronic drug exposure recruits brain stress systems — particularly corticotropin-releasing factor (CRF) signaling in the extended amygdala... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference6 A fundamentally different class of models formalizes the transition from positive to negative reinforcement that characterizes the progression from recreational drug use to dependence. The opponent process framework, most extensively developed by Koob and colleagues, posits that chronic drug exposure recruits brain stress systems — particularly corticotropin-releasing factor (CRF) signaling in the extended amygdala... -
1CitationThe systematic translation failures documented in {ref}
sec-translation-gap— from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...content/09_computational_models.md:line 4Open reference7 A fundamentally different class of models formalizes the transition from positive to negative reinforcement that characterizes the progression from recreational drug use to dependence. The opponent process framework, most extensively developed by Koob and colleagues, posits that chronic drug exposure recruits brain stress systems — particularly corticotropin-releasing factor (CRF) signaling in the extended amygdala... -
... 83 additional anchors in refs_json
References
- [Mollick2020a] “The systematic translation failures documented in {ref}`sec-translation-gap` — from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...”
- [Schultz2013] “The systematic translation failures documented in {ref}`sec-translation-gap` — from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...”
- [Koob2008a] “The systematic translation failures documented in {ref}`sec-translation-gap` — from candidate gene disappointments to the limited clinical predictive power of circuit-level findings — raise an urgent question: can formal computational models bridge the gap between preclinical mechanisms and human addiction? Computational approaches promise to impose mathematical discipline on the verbal theories that have guided the...”
- [Doll2012] “The landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...”
- [Smith2020a] “The landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...”
- [Deperrois2018] “The landscape of computational addiction models is dominated by reinforcement learning (RL) and temporal difference (TD) approaches, which account for the majority of formal models in the field [Mollick2020a, Schultz2013, Doll2012]. Opponent process and allostatic models form a second major cluster, followed by Bayesian and predictive coding frameworks, with circuit-level biophysical models representing the smallest...”
- [Gutkin2006] “The application of RL theory to addiction rests on the foundational observation that midbrain dopamine neurons encode reward prediction errors (RPEs) consistent with TD-learning algorithms [Schultz2013]. This mapping between a computational variable (the RPE) and a neural signal (phasic dopamine) provided a formal vocabulary for describing how drugs of abuse could hijack learning mechanisms: if addictive substances...”
- [Dezfouli2013] “The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits [Doll2012, Dezfouli2013, Kool2018]. Lesaint et al. (2014) showed that a combination of model-based and revised model...”
- [Kool2018] “The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits [Doll2012, Dezfouli2013, Kool2018]. Lesaint et al. (2014) showed that a combination of model-based and revised model...”
- [Lesaint2014] “The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits [Doll2012, Dezfouli2013, Kool2018]. Lesaint et al. (2014) showed that a combination of model-based and revised model...”
- [Fiore2017] “The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits [Doll2012, Dezfouli2013, Kool2018]. Lesaint et al. (2014) showed that a combination of model-based and revised model...”
- [Reiter2016] “The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits [Doll2012, Dezfouli2013, Kool2018]. Lesaint et al. (2014) showed that a combination of model-based and revised model...”
- [Wagner2021] “The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits [Doll2012, Dezfouli2013, Kool2018]. Lesaint et al. (2014) showed that a combination of model-based and revised model...”
- [Wiehler2019] “The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits [Doll2012, Dezfouli2013, Kool2018]. Lesaint et al. (2014) showed that a combination of model-based and revised model...”
- [Byrne2020] “The model-free versus model-based (MF/MB) distinction has become the most influential theoretical axis within this framework. Addiction, on the standard account, reflects a pathological shift from flexible, goal-directed (model-based) decision-making toward rigid, stimulus-response (model-free) habits [Doll2012, Dezfouli2013, Kool2018]. Lesaint et al. (2014) showed that a combination of model-based and revised model...”
- [Morris2019a] “Yet the MF/MB framework faces a fundamental challenge. Morris and Cushman (2019) argued that much compulsive drug-seeking behavior is not truly habitual but rather represents goal-directed pursuit of drug rewards — behavior that appears inflexible because the goal itself (obtaining the drug) dominates decision-making, not because the underlying computational architecture has shifted [Morris2019a]. This is not merely...”
- [Byrne2021] “Yet the MF/MB framework faces a fundamental challenge. Morris and Cushman (2019) argued that much compulsive drug-seeking behavior is not truly habitual but rather represents goal-directed pursuit of drug rewards — behavior that appears inflexible because the goal itself (obtaining the drug) dominates decision-making, not because the underlying computational architecture has shifted [Morris2019a]. This is not merely...”
- [Morris2019b] “Yet the MF/MB framework faces a fundamental challenge. Morris and Cushman (2019) argued that much compulsive drug-seeking behavior is not truly habitual but rather represents goal-directed pursuit of drug rewards — behavior that appears inflexible because the goal itself (obtaining the drug) dominates decision-making, not because the underlying computational architecture has shifted [Morris2019a]. This is not merely...”
- [Russek2016] “Several recent developments have enriched the RL landscape beyond the simple MF/MB dichotomy. Russek et al. (2016) introduced the successor representation (SR) as a bridge between model-free and model-based systems, demonstrating that the SR — a predictive state representation combined with TD learning — can produce a subset of behaviors associated with model-based learning while requiring less computation than full...”
- [Akam2020] “Akam and Walton (2020) reported findings unexpected under the standard RPE account: dopamine signals in certain contexts appear to carry information beyond simple model-free prediction errors, suggesting dopamine's involvement in model-based RL through mechanisms where dopamine neurons encode either state prediction errors or a combination of reward and state prediction information [Akam2020, Akam2021]. Shouval et a...”
- [Akam2021] “Akam and Walton (2020) reported findings unexpected under the standard RPE account: dopamine signals in certain contexts appear to carry information beyond simple model-free prediction errors, suggesting dopamine's involvement in model-based RL through mechanisms where dopamine neurons encode either state prediction errors or a combination of reward and state prediction information [Akam2020, Akam2021]. Shouval et a...”
- [Shouval2023] “Akam and Walton (2020) reported findings unexpected under the standard RPE account: dopamine signals in certain contexts appear to carry information beyond simple model-free prediction errors, suggesting dopamine's involvement in model-based RL through mechanisms where dopamine neurons encode either state prediction errors or a combination of reward and state prediction information [Akam2020, Akam2021]. Shouval et a...”
- [Cone2022] “Akam and Walton (2020) reported findings unexpected under the standard RPE account: dopamine signals in certain contexts appear to carry information beyond simple model-free prediction errors, suggesting dopamine's involvement in model-based RL through mechanisms where dopamine neurons encode either state prediction errors or a combination of reward and state prediction information [Akam2020, Akam2021]. Shouval et a...”
- [Cockburn2010] “The application of RL models to clinical populations beyond substance use disorders has produced a range of findings. Cockburn and Holroyd (2010) explored how asymmetric dopamine-mediated learning could account for features of ADHD [Cockburn2010]. Li et al. (2014) found that individuals with psychosis updated reward values more rapidly, consistent with elevated prediction error signaling [Li2014]. Balasubramani et a...”
- [Li2014] “The application of RL models to clinical populations beyond substance use disorders has produced a range of findings. Cockburn and Holroyd (2010) explored how asymmetric dopamine-mediated learning could account for features of ADHD [Cockburn2010]. Li et al. (2014) found that individuals with psychosis updated reward values more rapidly, consistent with elevated prediction error signaling [Li2014]. Balasubramani et a...”
- [Balasubramani2015] “The application of RL models to clinical populations beyond substance use disorders has produced a range of findings. Cockburn and Holroyd (2010) explored how asymmetric dopamine-mediated learning could account for features of ADHD [Cockburn2010]. Li et al. (2014) found that individuals with psychosis updated reward values more rapidly, consistent with elevated prediction error signaling [Li2014]. Balasubramani et a...”
- [Koob2008b] “A fundamentally different class of models formalizes the transition from positive to negative reinforcement that characterizes the progression from recreational drug use to dependence. The opponent process framework, most extensively developed by Koob and colleagues, posits that chronic drug exposure recruits brain stress systems — particularly corticotropin-releasing factor (CRF) signaling in the extended amygdala...”
- [Koob2013] “A fundamentally different class of models formalizes the transition from positive to negative reinforcement that characterizes the progression from recreational drug use to dependence. The opponent process framework, most extensively developed by Koob and colleagues, posits that chronic drug exposure recruits brain stress systems — particularly corticotropin-releasing factor (CRF) signaling in the extended amygdala...”
- [Koob2015] “A fundamentally different class of models formalizes the transition from positive to negative reinforcement that characterizes the progression from recreational drug use to dependence. The opponent process framework, most extensively developed by Koob and colleagues, posits that chronic drug exposure recruits brain stress systems — particularly corticotropin-releasing factor (CRF) signaling in the extended amygdala...”
- [George2012] “A fundamentally different class of models formalizes the transition from positive to negative reinforcement that characterizes the progression from recreational drug use to dependence. The opponent process framework, most extensively developed by Koob and colleagues, posits that chronic drug exposure recruits brain stress systems — particularly corticotropin-releasing factor (CRF) signaling in the extended amygdala...”
Sister wikis (recently updated · no domain on this page)
- Agent Recipe: AI-for-Biology Closed-Loop with Reviewer Handoffs and Eval Contracts
- Agent Recipe: AI-for-Biology Closed-Loop with Reviewer Handoffs and Eval Contracts
- test
- JGBO-I27: Top 10 GBO Questions for Prioritization
- JGBO-I27: Top 10 GBO Questions for Prioritization
- Design Brief: Beta-test Evaluation Protocol for SciDEX v2 Design Trajectories
- Andy — Showcase Findings (auto-curated)
- Kris — Showcase Findings (auto-curated)
Recent activity here
No recent events touching this page.