LLMs exhibit Bayesian-like hypothesis updating with strong-sampling bias and an evaluation-generation gap but generalize poorly outside observed data.
Synaptic reverberation underlying mnemonic persistent activity.Trends in Neurosciences, 24(8):455–463
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
Recurrent networks driven by low-dimensional sensory dynamics generically embed those dynamics as smooth internal manifolds, with prediction accuracy forcing state separation up to a resolution limit set by prediction error.
citing papers explorer
-
Hypothesis generation and updating in large language models
LLMs exhibit Bayesian-like hypothesis updating with strong-sampling bias and an evaluation-generation gap but generalize poorly outside observed data.
-
Embedding of Low-Dimensional Sensory Dynamics in Recurrent Networks: Implications for the Geometry of Neural Representation
Recurrent networks driven by low-dimensional sensory dynamics generically embed those dynamics as smooth internal manifolds, with prediction accuracy forcing state separation up to a resolution limit set by prediction error.