SAE Feature Interpretability Improves Across Layers
Sauers_ · x · 2026-07-14
This repost discusses how jlens is used to describe SAE features, essentially verbalizing highly correlated features like "motor neurons."
By comparing the consistency between jlens-generated feature descriptions and Neuronpedia labels, the author observed that this "verbalization" capability improves with depth. In Qwen3-4b, a significant shift occurs around layer 22. The original post also notes a similar trend in the raw Jacobian results, where later layers exhibit higher Jacobian gains.
More from Research
- KANLab: an interactive browser playground for Kolmogorov-Arnold Networks — Lukas Niehaus · 2026-10-27
- CMU's LLM2Cert synthesizes stability certificates verified via SMT and Lean — CarnegieMellonU · 2026-10-27
- NVIDIA open-sources FLORA, a robotics data engine ~125x cheaper than GPT-6 Astra — nvidia · 2026-10-27
- RSNA 2026 Deep Learning Labs return: eight hands-on AI classes over four days in Chicago — PRouzrokh · 2026-10-27
- Steering two orthogonal concept neurons makes a model compose "gnome" — Sauers_ · 2026-10-27
- Paper: data schedules and learning rate schedules should be designed together — alifcoder · 2026-10-27