Qwen3-8B has an internal direction for 'this body part is mine'
repligate · x · 2026-11-11
At layer 24 of Qwen3-8B, euagaric built an internal direction encoding 'this body part is mine' from contrastive touch sentences ('your' vs 'the woman's'), then scored new sentences on it. The direction tracks whose body is touched, not who does the touching, and persists after the model agrees to notice body sensations. In plain chat, the same direction mostly tracks the word 'you'. Edge cases (mirrors, robots, lost hands) are charted.
More from Research
- Scandinavian Journal of Statistics editors reflect on publishing in the age of generative AI — michaelchchoi · 2026-11-11
- Proving LLM backdoor-safety without revealing weights: ZKML over formal methods — DimitrisPapail · 2026-11-11
- Claude Opus 5.5 runs an architecture search and finds AttnRes-Canon, a new top Transformer design — SonglinYang4 · 2026-11-11
- 5 research habits worth knowing: read papers backwards, reproduce first, wire coding agents via MCP — cneuralnetwork · 2026-11-11
- Running Qwen2.5-Coder-3B on Android: Jupyter + Pydantic Evals on-device eval pipeline — alex_b3b · 2026-11-11
- cloneofsimo releases a video explaining determinantal point processes — cloneofsimo · 2026-11-11