List your company

General Mathematics Reasoning on College-math, Math-OAI, Minerva-math (test)

54.1AccuracyRej@4
0.74814.59928.4542.301Feb 12, 2026

Evaluation results

MethodLinks
Rej@4Base Model=Qwen2.5-7B-instruct, Budget (FLOPs)=60.19 x 10^182026.0254.1
SimPOBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=77.88 x 10^182026.0254.1
HD+SFTBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=44.70 x 10^182026.0253.9
CPOBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=77.88 x 10^182026.0253.7
OriginBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=02026.0253.5
DPOBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=77.88 x 10^182026.0253.5
Rej@2Base Model=Qwen2.5-7B-instruct, Budget (FLOPs)=46.03 x 10^182026.0253.4
HD+IDFTBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=44.70 x 10^182026.0253.2
RPOBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=77.88 x 10^182026.0252.1
HD+IDFTBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=19.86 x 10^182026.0241.9
Rej@4Base Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=35.81 x 10^182026.0241.6
SimPOBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=40.39 x 10^182026.0241.6
Rej@2Base Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=25.40 x 10^182026.0238.9
OriginBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=02026.0238.1
HD+SFTBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=19.86 x 10^182026.0237.5
RPOBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=40.39 x 10^182026.0236.1
DPOBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=40.39 x 10^182026.0235.7
CPOBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=40.39 x 10^182026.0231.7
EAFTBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=31.87 x 10^182026.0227
SFTBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=31.87 x 10^182026.0225.3
DFTBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=14.99 x 10^182026.0224.8
DFTBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=31.87 x 10^182026.0222.8
EAFTBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=14.99 x 10^182026.0221.4
SFTBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=14.99 x 10^182026.022.8