DailyPapers(@HuggingPapers):Do LLMs actually think in silence? A new paper formalizes four axioms — Causality, Minimality, Separability, and Stability — to audit latent thought representations. No open-weight LLM satisfies all four. Most representations encode nothing beyond the input embedding itself.

DailyPapers

@HuggingPapers

Tweeting interesting papers submitted at Submit your own at and link models/datasets/demos to it!

Joined March 2025

4 Following 20.1K Followers

DailyPapers@HuggingPapers

2026.06.29 17:04

Do LLMs actually think in silence? A new paper formalizes four axioms — Causality, Minimality, Separability, and Stability — to audit latent thought representations. No open-weight LLM satisfies all four. Most representations encode nothing beyond the input embedding itself.