On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,742 papers · 148 categories

Trend · papers per month

count % of corpus

4 results for “Polysemanticity”

Taming Polysemanticity in LLMs: Provable Feature Recovery via Sparse Autoencoderscs.LG

New SAE algorithm proves feature recovery for LLMs with theoretical guarantees.

problem Achieving interpretable features in large language models (LLMs).

method Proposed a statistical framework and bias adaptation technique for sparse autoencoders (SAEs).

result Proved correct recovery of all monosemantic features under specific data sampling.

Large Language Models Sparse Autoencoders Feature Recovery Polysemanticity Bias Adaptation

2025-06-16abs ↗pdf ↗

Effects of sparsity and superposition on loss in simple autoencoderscs.LG

Superposition in autoencoders leads to loss in simple models.

problem Mechanistic interpretability of neural networks

method Analyzing mathematical basis for superposition and providing bounds for reconstruction loss

result Upper and lower bounds for L2 reconstruction loss in the sparse regime

superposition autoencoder interpretability loss neural networks

2026-06-16abs ↗pdf ↗

Compositional Explanations of Neuronscs.LG

Describes explaining neurons in deep representations using compositional logical concepts.

problem Interpreting neuron behavior in deep neural networks.

method Identifying compositional logical concepts that closely approximate neuron behavior.

result Compositional explanations provide insights into model performance and allow for adversarial example creation.

neural networks interpretability deep learning neuron explanations adversarial examples

2020-06-24abs ↗pdf ↗

Feature Starvation as Geometric Instability in Sparse Autoencoderscs.LG

AEN-SAEs address feature starvation in sparse autoencoders by stabilizing the geometric alignment of sparse coding.

problem Feature starvation in sparse autoencoders, leading to unstable and misaligned representations.

method Adaptive Elastic Net SAEs (AEN-SAEs) combine

\ell_2

and

\ell_1

terms to stabilize the sparse coding map and control feature interactions.

result AEN-SAEs mitigate feature starvation without heuristic resampling, maintaining competitive reconstruction abilities.

sparse autoencoders feature starvation geometric instability sparse regression Lipschitz continuity

2026-05-06abs ↗pdf ↗