LLMs exhibit Bayesian-like hypothesis updating with strong-sampling bias and an evaluation-generation gap but generalize poorly outside observed data.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 3years
2026 3verdicts
UNVERDICTED 3roles
background 1polarities
background 1representative citing papers
Fourier embeddings create periodic vector representations that support Dirichlet and periodic Gaussian kernels within Spatial Semantic Pointers.
Radial basis kernels are realizable in spatial semantic pointers via distributed Fourier embeddings, with grid cell-like codes being capable and optimal.
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.
-
On periodic distributed representations using Fourier embeddings
Fourier embeddings create periodic vector representations that support Dirichlet and periodic Gaussian kernels within Spatial Semantic Pointers.
-
Neurally-plausible radial basis kernels using distributed Fourier embeddings
Radial basis kernels are realizable in spatial semantic pointers via distributed Fourier embeddings, with grid cell-like codes being capable and optimal.