2 papers
cs.LG2025
Mechanistic Permutability: Match Features Across Layers
Nikita Balagansky, Ian Maksimov, Daniil Gavrilov
Understanding how features evolve across layers in deep neural networks is a fundamental challenge in mechanistic interpretability, particularly due to polysemanticity and feature…
cs.LG2025
Learn Your Reference Model for Real Good Alignment
Alexey Gorbatovski, Boris Shaposhnikov, Alexey Malakhov +5
Despite the fact that offline methods for Large Language Models (LLMs) alignment do not require a direct reward model, they remain susceptible to overoptimization. This issue arise…