List your company

Reasoning on GPQA-D (avg@k, Token usage)

77.46Avg@kType IV (STOP)
30.514442.702254.8967.0778Apr 17, 2026

Evaluation results

MethodLinks
Type IV (STOP)Model=GPT-OSS-20B2026.0477.46143.4-48.26
Type IIIModel=GPT-OSS-20B2026.0468.43145.9-47.36
Type IModel=GPT-OSS-20B2026.0466.41151.8-45.24
No pruning (Baseline)Model=GPT-OSS-20B2026.0465.55277.2—
Type IV (STOP)Model=DS-Qwen-3-8B2026.0463.32193.5-70.35
Type IIIModel=DS-Qwen-3-8B2026.0458.78196.6-69.86
Type IModel=DS-Qwen-3-8B2026.0457.07201-69.2
No pruning (Baseline)Model=DS-Qwen-3-8B2026.0456.87652.6—
Type IV (STOP)Model=DS-Qwen-2.5-7B2026.0455.75165.9-62.61
Type IV (STOP)Model=DS-Qwen-2.5-1.5B2026.0448.42179.4-67.43
Type IModel=DS-Qwen-2.5-7B2026.0447.09173.6-60.89
Type IIIModel=DS-Qwen-2.5-7B2026.0446.02166.7-62.43
No pruning (Baseline)Model=DS-Qwen-2.5-7B2026.0445.95443.8—
Type IIIModel=DS-Qwen-2.5-1.5B2026.0434.98180.4-67.25
No pruning (Baseline)Model=DS-Qwen-2.5-1.5B2026.0433.08550.9—
Type IModel=DS-Qwen-2.5-1.5B2026.0432.32187.1-66.03