Code-as-World: Agentic Discovery of Executable World Representations for Physical Reasoning
MirroS-Lab · hf · 2026-08-31
Code-as-World represents physical environments as executable code to enable quantitative reasoning and scalable supervision for vision-language models, enhancing understanding and manipulation in physical scenes.
More from Research
- How to Build a Diffusion Language Model: A Complete Guide — zainhas · 2026-08-31
- PILOT framework enables live self-improvement for long-horizon agents — rohanpaul_ai · 2026-08-31
- Simulation shows high LLM judge reliability increases false positive risk — IanArawjo · 2026-08-31
- Opinion: Discard "Persona", Rely on Pretraining Reasoning — voooooogel · 2026-08-31
- Sori-1B: Audio-Grounded LM Trained From Scratch With No Text-Only Pretraining — Balance- · 2026-08-31
- TEMPO enables models to explore, hypothesize, and learn lifelong — heyshrutimishra · 2026-08-31