List your company

Closed-world plan-synthesis on ALFWorld unseen (val)

100Success Rate (SR)Fast Downward (upper bound)
77.53683.36889.295.032Jul 7, 2026

Evaluation results

MethodLinks
Fast Downward (upper bound)2026.07100—
GPT-5.5 (LLM+VAL loop)Backbone=GPT-5.52026.071002,770,945
EvoPlanBackbone=Qwen3-32B2026.0790.310,349,634
EvoPlanBackbone=gpt-oss-120b2026.0778.417,071,316