List your company

Competition Mathematics Reasoning on AIME24, AMC23, and Olympiadbench (test)

34.6AccuracyHD+IDFT
9.74416.19722.6529.103Feb 12, 2026

Evaluation results

MethodLinks
HD+IDFTBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=44.70 x 10^182026.0234.6
HD+SFTBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=44.70 x 10^182026.0234.1
SimPOBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=77.88 x 10^182026.0233.9
OriginBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=02026.0233.6
DPOBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=77.88 x 10^182026.0233.6
CPOBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=77.88 x 10^182026.0233.5
Rej@2Base Model=Qwen2.5-7B-instruct, Budget (FLOPs)=46.03 x 10^182026.0232.9
Rej@4Base Model=Qwen2.5-7B-instruct, Budget (FLOPs)=60.19 x 10^182026.0232.7
RPOBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=77.88 x 10^182026.0232.4
SimPOBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=40.39 x 10^182026.0224.6
Rej@4Base Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=35.81 x 10^182026.0222
HD+IDFTBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=19.86 x 10^182026.0219.8
Rej@2Base Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=25.40 x 10^182026.0218.1
HD+SFTBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=19.86 x 10^182026.0217.8
DPOBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=40.39 x 10^182026.0217.4
EAFTBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=31.87 x 10^182026.0216.9
RPOBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=40.39 x 10^182026.0216.8
OriginBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=02026.0216.5
CPOBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=40.39 x 10^182026.0216.2
SFTBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=31.87 x 10^182026.0215.1
SFTBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=14.99 x 10^182026.0214.4
DFTBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=14.99 x 10^182026.0211.9
EAFTBase Model=Phi4-mini-instruct (3.8B), Budget (FLOPs)=14.99 x 10^182026.0211.8
DFTBase Model=Qwen2.5-7B-instruct, Budget (FLOPs)=31.87 x 10^182026.0210.7