List your company

Scientific Reasoning on GPQA Diamond (Accuracy MV, Accuracy WMV, Required Tokens)

72.7Accuracy (MV)DeepConf
57.51661.45865.469.342Jun 30, 2026

Evaluation results

MethodLinks
DeepConfBackbone=Phi-4-RP-14B, branching factor n=322026.0672.772.349.3
Flex-Fork-thinkBackbone=Phi-4-RP-14B, branching factor n=32, seed path length l0=2,0482026.0670.763.237.8
DeepConfBackbone=DeepSeek-8B, branching factor n=322026.0667.366.373.8
ASCBackbone=DeepSeek-8B, branching factor n=322026.0667.1—42.1
Flex-Fork-thinkBackbone=Qwen3-8B, branching factor n=32, seed path length l0=2,0482026.0663.163.235
DeepConfBackbone=Qwen3-8B, branching factor n=322026.0662.862.364.2
Flex-Fork-thinkBackbone=DeepSeek-8B, branching factor n=32, seed path length l0=2,0482026.0662.163.134.5
ASCBackbone=Qwen3-8B, branching factor n=322026.0661.7—58.5
ASCBackbone=Phi-4-RP-14B, branching factor n=322026.0658.1—41.8