List your company

Sequential Embodied Task Performance on ALFWorld 140-episode

97.9Average Reward (%)GoS
45.06858.78472.586.216Apr 7, 2026

Evaluation results

MethodLinks
GoSModel=Claude Sonnet 4.52026.0497.927,21549.2
Vector SkillsModel=Claude Sonnet 4.52026.0493.628,40737.8
GoSModel=GPT-5.2 Codex2026.0493.646,46264.7
Vector SkillsModel=GPT-5.2 Codex2026.0492.934,43657
Vanilla SkillsModel=Claude Sonnet 4.52026.0489.31,524,40153.2
Vanilla SkillsModel=GPT-5.2 Codex2026.0489.31,435,61483.3
GoSModel=MiniMax M2.72026.0454.365,22768.8
Vector SkillsModel=MiniMax M2.72026.0450.766,10973.4
Vanilla SkillsModel=MiniMax M2.72026.0447.12,184,82388.6