NTU's Trace2Env Builds Agent Environment Simulations from Historical Traces
nanyang-technological-university-singapore · hf · 2026-10-09
NTU introduces Trace2Env, an agentic language world modeling paradigm where a world model agent serves as the environment for task agents. When the original system is unavailable but interaction traces exist, Trace2Env reconstructs them into a reusable worldbook of schemas, evidence, and behavioral knowledge, consulted at runtime alongside persistent episodic state. Across nine environments it improves next-observation fidelity and long-horizon consistency over prompt-based LWMs, with task agent actions remaining valid more often when replayed in real environments.
More from coding & agent
- IBM's RIT-RAG induces document sub-trees from retrieved chunks, lifting RAG accuracy by up to 11.4 points — _reachsumit · 2026-10-09
- Reddit thread rounds up open-source coding agents like MonkeyCode and OpenHands for local use — Good_Full_Tmes · 2026-10-09
- OpenAI's 10,000-agent, 130B-token run pushed slime v0.4.0 to rethink RL infrastructure scale — teortaxesTex · 2026-10-09
- Voice agent stack breakdown: Twilio, Deepgram, Cresta and ElevenLabs with a 300ms latency rule — schwentker · 2026-10-09
- Refund tools as verbs, not pens: how 'Veronica' stops LLM agents from inventing payees — schwentker · 2026-10-09
- Hackathon build 'Veronica': a voice agent that handles 'you charged me 10x' calls — schwentker · 2026-10-09