List your company

Reasoning on BBEH (Accuracy, Delta Avg)

81.2AccuracyCoT2-Meta
63.83268.34172.8577.359Mar 30, 2026

Evaluation results

MethodLinks
CoT2-MetaBackbone=DeepSeek-V3.2, Strategy=Ours (CoT2-Meta)2026.0381.210.5
Vanilla ToTBackbone=DeepSeek-V3.2, Strategy=Vanilla ToT2026.0377.85.7
Best-of-16Backbone=DeepSeek-V3.2, Strategy=Best-of-162026.0375.93
CoT2-MetaBackbone=Claude-4.5, Strategy=Ours (CoT2-Meta)2026.0375.814.5
Greedy CoTBackbone=DeepSeek-V3.2, Strategy=Greedy CoT2026.0374.4—
CoT2-MetaBackbone=Qwen2.5-VL-7B, Strategy=Ours (CoT2-Meta)2026.0372.512.2
Vanilla ToTBackbone=Claude-4.5, Strategy=Vanilla ToT2026.0371.28.3
Vanilla ToTBackbone=Qwen2.5-VL-7B, Strategy=Vanilla ToT2026.0369.16.4
Best-of-16Backbone=Claude-4.5, Strategy=Best-of-162026.0368.94.8
Best-of-16Backbone=Qwen2.5-VL-7B, Strategy=Best-of-162026.0366.83.4
Greedy CoTBackbone=Claude-4.5, Strategy=Greedy CoT2026.0365.4—
Greedy CoTBackbone=Qwen2.5-VL-7B, Strategy=Greedy CoT2026.0364.5—