List your company

Instruction Following on ALFWorld Hard

20.66Success Rate (SR)TCOD (B2F)
0.03685.390910.74516.0991Apr 27, 2026

Evaluation results

MethodLinks
TCOD (B2F)Student Model=Qwen2.5-7B2026.0420.6627.07
TCOD (F2B)Student Model=Qwen2.5-7B2026.0418.1827.37
TCOD (B2F)Student Model=Qwen2.5-3B2026.0413.2228.16
Vanilla OPDStudent Model=Qwen2.5-7B2026.0413.2227.89
Vanilla OPDStudent Model=Qwen2.5-3B2026.0410.7428.64
TCOD (F2B)Student Model=Qwen2.5-3B2026.049.9228.57
SFTStudent Model=Qwen2.5-7B2026.048.2628.73
Qwen2.5-7B-RL (Teacher)Model Type=Teacher2026.046.6127.31
SFTStudent Model=Qwen2.5-3B2026.044.9629.12
Zero-ShotStudent Model=Qwen2.5-7B2026.041.6529.77
Zero-ShotStudent Model=Qwen2.5-3B2026.040.8329.88