- claim_text
A recurrent neural network optimized to predict upcoming sensory inputs from natural visual stimuli spontaneously develops the functional-connectivity rules observed empirically in mouse V1, including like-to-like excitatory connectivity, the relationship between connectivity and orientation/direction tuning, and the response-similarity gradient with connection strength — suggesting that temporal prediction may be an underlying organizational principle of V1 lateral connectivity.
- raw_fields
{
"n": 0,
"doi": "10.1016/j.cub.2024.11.073",
"claim": "A recurrent neural network optimized to predict upcoming sensory inputs from natural visual stimuli spontaneously develops the functional-connectivity rules observed empirically in mouse V1, including like-to-like excitatory connectivity, the relationship between connectivity and orientation/direction tuning, and the response-similarity gradient with connection strength — suggesting that temporal prediction may be an underlying organizational principle of V1 lateral connectivity.",
"cite_key": "KlavinskisWhiting2025",
"evidence": "Training RNNs on a temporal-prediction objective with natural visual movies and comparing emergent connectivity statistics to published mouse-V1 paired-recording datasets.",
"effect_size": "qualitative — temporal-prediction objective recapitulates multiple empirical V1 wiring biases",
"text_access": "fulltext",
"study_system": "computational model with mouse V1 empirical-data validation",
"argument_role": "supporting",
"replication_status": "single_study",
"claim_source_sentence": "the functional specificity of V1 connections emerges naturally in a recurrent neural network optimized to predict upcoming sensory inputs for natural visual stimuli. This temporal prediction model reproduces the complex relationships between the connectivity of V1 neurons and their orientation and direction preferences",
"source_provenance_status": "ok",
"replication_evidence_dois": [],
"effect_size_source_sentence": "the functional specificity of V1 connections emerges naturally in a recurrent neural network optimized to predict upcoming sensory inputs for natural visual stimuli"
}- source_refs
[
"paper:paper-d66965754b34"
]
- source_span
the functional specificity of V1 connections emerges naturally in a recurrent neural network optimized to predict upcoming sensory inputs for natural visual stimuli. This temporal prediction model reproduces the complex relationships between the connectivity of V1 neurons and their orientation and direction preferences
- evidence_refs
[
{
"ref": "paper:paper-d66965754b34"
}
]- source_policy
{
"mode": "public_source_pointer_with_short_context",
"notes": [
"Local review repositories are read-only inputs.",
"SciDEX stores paper metadata, structured evidence, file pointers, and short citation contexts; it does not copy full review prose."
],
"source_commit_sha": "79ce062d54a924ce05953ec90aa9d26044d2b48f",
"source_repository_url": "https://github.com/AllenNeuralDynamics/ComputationalReviewRecurrence"
}