List your company

Mathematical Reasoning on Minerva (Pass@1 Accuracy, Avg Length)

54.53Pass@1 AccuracyPass@k†
20.802829.558938.31547.0711Mar 18, 2026Mar 29, 2026Apr 9, 2026Apr 20, 2026May 1, 2026May 12, 2026May 24, 2026

Evaluation results

MethodLinks
Pass@k†Backbone=Phi-4-mini-Instruct, N=322026.0554.5317,119
Pass@k†Backbone=Qwen2.5-7B-Instruct, N=322026.0552.3319,765
ExpThinkBackbone=Qwen3-8B2026.0545.5850.4
VanillaBackbone=Qwen3-8B2026.0544.17,210
LaserBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=7B2026.0343.91,826
SPSBackbone=Qwen2.5-7B-Instruct, N=82026.0543.269,387
LC-R1Base Model=DeepSeek-R1-Distill-Qwen, Model Scale=7B2026.0342.81,298.8
TRiMSBackbone=Qwen3-4B2026.0341.9783
PB-SMCBackbone=Qwen2.5-7B-Instruct, N=82026.0541.7913,216
VanillaBackbone=Qwen3-4B2026.0341.56,673.1
JETBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0540.43,333.9
SMCBackbone=Qwen2.5-7B-Instruct, N=322026.0540.3221,642
Beam SearchBackbone=Qwen2.5-7B-Instruct, N=322026.0540.223,900
DVTSBackbone=Qwen2.5-7B-Instruct, N=322026.0539.2223,549
TRiMSBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=7B2026.0339696.8
Self-ConsistencyBackbone=Qwen2.5-7B-Instruct, N=322026.0538.8519,765
ExpThinkBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0538.61,182.7
Best-of-NBackbone=Qwen2.5-7B-Instruct, N=322026.0538.4819,765
PB-SMCBackbone=Phi-4-mini-Instruct, N=82026.0537.8713,198
AutoThinkBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0537.82,617.4
JETBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=7B2026.0337.51,786.1
VanillaBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0537.55,017.3
LaserBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0537.12,050.9
MCTSBackbone=Qwen2.5-7B-Instruct, N=322026.0536.2725,208
SPSBackbone=Phi-4-mini-Instruct, N=82026.0535.429,832
AutoThinkBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=7B2026.0335.32,512.2
AdaptThinkBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=7B2026.0334.92,362
Best-of-NBackbone=Phi-4-mini-Instruct, N=322026.0534.817,119
Base ModelBackbone=Qwen2.5-7B-Instruct, N=12026.0534.68617
SMCBackbone=Phi-4-mini-Instruct, N=322026.0534.4424,907
VanillaBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=7B2026.0334.24,907.5
DVTSBackbone=Phi-4-mini-Instruct, N=322026.0533.8222,140
AdaptThinkBackbone=DeepSeek-R1-Distill-Qwen-7B2026.0533.82,645.4
Self-ConsistencyBackbone=Phi-4-mini-Instruct, N=322026.0533.717,119
Beam SearchBackbone=Phi-4-mini-Instruct, N=322026.0533.4623,768
LaserBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=1.5B2026.0333.13,218.3
MCTSBackbone=Phi-4-mini-Instruct, N=322026.0532.7523,108
LC-R1Backbone=DeepSeek-R1-Distill-Qwen-7B2026.0532.41,734.4
AutoThinkBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=1.5B2026.03293,031.5
ExpThinkBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.05291,761
LaserBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0528.73,185.5
TRiMSBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=1.5B2026.0328.3958.3
VanillaBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0526.86,347.1
LC-R1Base Model=DeepSeek-R1-Distill-Qwen, Model Scale=1.5B2026.0326.53,827.3
LC-R1Backbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0526.12,383.3
VanillaBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=1.5B2026.0325.76,284.1
AdaptThinkBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=1.5B2026.0325.51,876.8
JETBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0525.43,988.1
AutoThinkBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0524.33,201.2
JETBase Model=DeepSeek-R1-Distill-Qwen, Model Scale=1.5B2026.0323.91,839.1
Base ModelBackbone=Phi-4-mini-Instruct, N=12026.0523.9535
AdaptThinkBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.0522.11,968.5