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

Completion on HOWTOBENCH

0.91Pearson Correlation (ρ)

Elaborated Rubric

Updated 1mo ago

Evaluation Results

Method	Links
Elaborated Rubric 2026.04		0.91	0.83	0.92
Average scoring (ToW w/o plan) 2026.04		0.9	0.61	0.82
Elaborated Rubric + SC 2026.04		0.9	0.83	0.9
Elaborated Rubric + SC 2026.04		0.9	0.83	0.9
ToW 2026.04		0.89	0.78	0.88
Elaborated Rubric 2026.04		0.89	0.83	0.9
ROUGE-L 2026.04		0.87	0.67	0.75
ToW 2026.04		0.87	0.67	0.78
Elaborated Rubric 2026.04		0.86	0.56	0.7
BLEU-1 2026.04		0.85	0.67	0.8
Average scoring (ToW w/o plan) 2026.04		0.85	0.56	0.67
Elaborated Rubric + SC 2026.04		0.85	0.56	0.7
Elaborated Rubric + SC 2026.04		0.85	0.56	0.7
Auto-Plan 2026.04		0.85	0.72	0.82
Elaborated Rubric 2026.04		0.84	0.56	0.72
Auto-Plan + SC 2026.04		0.84	0.67	0.85
Auto-Plan + SC 2026.04		0.84	0.67	0.77
Auto-Plan 2026.04		0.83	0.61	0.73
Auto-Plan 2026.04		0.73	0.56	0.72
Auto-Plan + SC 2026.04		0.73	0.42	0.57
Auto-Plan + SC 2026.04		0.73	0.39	0.47
BLEU-1 2026.04		0.7	0.5	0.62
Auto-Plan 2026.04		0.63	0.4	0.49
ROUGE-L 2026.04		0.46	0.22	0.32
BLEU-rt 2026.04		0.19	0.06	0.15
BLEU-rt 2026.04		-0.45	-0.22	-0.27