Share your thoughts, 1 month free Claude Pro on usSee more

Interactive Reasoning on ALFWorld

47.9Average Reasoning Length (tokens)

KD (1.3B)

Updated 1mo ago

Evaluation Results

Method	Links
KD (1.3B) 2025.05		47.9
SeqKD (1.3B) 2025.05		46.8
Token-level-1.3B 2025.05		45.7
KD (2.7B) 2025.05		44.3
SeqKD (2.7B) 2025.05		43.2
KD (7B) 2025.05		43
KD (6.7B) 2025.05		42.6
Token-level-2.7B 2025.05		42.5
SeqKD (7B) 2025.05		42
SeqKD (6.7B) 2025.05		41.7
Ours (OPT-1.3B) 2025.05		41.2
Token-level-7B 2025.05		41.1
Token-level-OPT-6.7B 2025.05		40.8
Ours (OPT-2.7B) 2025.05		39.4
Teacher (OPT-13B) 2025.05		38.2
Ours (LLaMA-7B) 2025.05		38.2
Ours (OPT-6.7B) 2025.05		38
Teacher (LLaMA-13B) 2025.05		37.5