Represented but Not Used: Value-Policy Dissociation in LLM Agents
arXiv:2611.10457v1 Announce Type: new Abstract: Large language model (LLM) agents need to choose actions whose consequences unfold over multiple future steps, yet poor decisions do not reveal whether the model failed to represent those consequences or failed to use them. Mechanistic interpretability has studied…
Excerpt provided by arXiv cs.AI via its public feed. Full reporting belongs to the original publisher.
Read the full story on arXiv cs.AI
Related coverage
The Decoder · 42m ago
Gemini 4 Carbon arrives just in time to feel like yesterdays news in the fastest model cycle yet
Ars Technica AI · 4h ago
DeepMind reveals Gemini Robotics 2.5 and open safety benchmark "ASIMOV-jailbreak"
Google DeepMind · 6h ago
Introducing AlphaProof Code: a breakthrough in verifying complex software at depth
OpenAI · 10h ago
Closing the AI guidance gap in education