Leon

Training-Free Self-Improvement Agents Part 1
11 hrs ago • Leon
7:47

September 2025

LLM hallucinations are compression failures, and we can detect them.
We previously argued that transformers behave Bayesian only on average over orderings, we now show when the information budget is too small, especially…
Sep 15, 2025 • Leon
© 2026 Leon · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture