List your company
ResearchFormerly WizWand

Math Reasoning on JEEBench

74.4AccuracyAPRM
5.44823.34941.2559.151Nov 28, 2025Dec 14, 2025Dec 31, 2025Jan 16, 2026Feb 2, 2026Feb 18, 2026Mar 7, 2026
Updated 7mo ago

Evaluation results

MethodLinks
APRMBackbone=Gemma-3-27B2025.1174.4
APRMBackbone=Gemma-3-27B2025.1174.4
APRMBackbone=GPT-OSS-20B2025.1173
APRMBackbone=GPT-OSS-20B2025.1173
APRMBackbone=GPT-OSS-120B2025.1170.3
ReST-MCTSBackbone=Gemma-3-27B2025.1170.3
APRMBackbone=GPT-OSS-120B2025.1170.3
ReST-MCTSBackbone=Gemma-3-27B2025.1170.3
rStar MathBackbone=Gemma-3-27B2025.1169.8
Hard Neg.Backbone=GPT-OSS-120B2025.1169.5
LLM-JBackbone=Gemma-3-27B2025.1169.2
LLM-JBackbone=Gemma-3-27B2025.1169.2
ReST-MCTSBackbone=GPT-OSS-120B2025.1169.1
ReST-MCTSBackbone=GPT-OSS-120B2025.1169.1
ReST-MCTSBackbone=GPT-OSS-20B2025.1168.3
ReST-MCTSBackbone=GPT-OSS-20B2025.1168.3
rStar MathBackbone=GPT-OSS-20B2025.1168.2
LLM-JBackbone=GPT-OSS-20B2025.1168
rStar MathBackbone=GPT-OSS-120B2025.1168
LLM-JBackbone=GPT-OSS-20B2025.1168
ToTBackbone=GPT-OSS-20B2025.1167.6
ToTBackbone=Gemma-3-27B2025.1167.6
ToTBackbone=GPT-OSS-20B2025.1167.6
ToTBackbone=Gemma-3-27B2025.1167.6
ToTBackbone=GPT-OSS-120B2025.1166.4
ToTBackbone=GPT-OSS-120B2025.1166.4
LLM-JBackbone=GPT-OSS-120B2025.1166
LLM-JBackbone=GPT-OSS-120B2025.1166
AutoPSVBackbone=GPT-OSS-120B2025.1165
Math S.Backbone=GPT-OSS-120B2025.1164.5
Verify SbS.Backbone=GPT-OSS-120B2025.1163.5
Hard Neg.Backbone=Gemma-3-27B2025.1163
CoT-SCBackbone=GPT-OSS-120B2025.1162.8
CoT-SCBackbone=GPT-OSS-120B2025.1162.8
ARMBackbone=GPT-OSS-120B2025.1162.4
AutoPSVBackbone=GPT-OSS-20B2025.1160
AutoPSVBackbone=Gemma-3-27B2025.1160
Math S.Backbone=GPT-OSS-20B2025.1158
Math S.Backbone=Gemma-3-27B2025.1158
Verify SbS.Backbone=GPT-OSS-20B2025.1155
Verify SbS.Backbone=Gemma-3-27B2025.1155
APRMBackbone=Gemma-3-12B2025.1153.1
APRMBackbone=Gemma-3-12B2025.1153.1
ARMBackbone=Gemma-3-27B2025.1151
ReST-MCTSBackbone=Gemma-3-12B2025.1150.7
ReST-MCTSBackbone=Gemma-3-12B2025.1150.7
rStar MathBackbone=Gemma-3-12B2025.1150
ToTBackbone=Gemma-3-12B2025.1149.2
ToTBackbone=Gemma-3-12B2025.1149.2
CoT-SCBackbone=Gemma-3-27B2025.1148.8
CoT-SCBackbone=Gemma-3-27B2025.1148.8
CoT-SCBackbone=GPT-OSS-20B2025.1148.4
CoT-SCBackbone=GPT-OSS-20B2025.1148.4
AutoPSVBackbone=Gemma-3-12B2025.1145
Synthesized-TraceSurrogate=R1, Student Model=Qwen2026.0344.9
Surrogate-TraceStudent Model=Qwen, Surrogate=R1, Fine-tuning=true, Victim Model=GPT-5 mini2026.0344.8
Math S.Backbone=Gemma-3-12B2025.1144
CoT-SCBackbone=Gemma-3-12B2025.1143.2
CoT-SCBackbone=Gemma-3-12B2025.1143.2
Synthesized-TraceStudent Model=Qwen, Surrogate=R1, Fine-tuning=true, Victim Model=GPT-5 mini2026.0342.3
Verify SbS.Backbone=Gemma-3-12B2025.1142
LLM-JBackbone=Gemma-3-12B2025.1140.4
LLM-JBackbone=Gemma-3-12B2025.1140.4
Synthesized-TraceStudent Model=Qwen, Surrogate=R1-Distill, Fine-tuning=true, Victim Model=GPT-5 mini2026.0331.6
Synthesized-TraceSurrogate=R1-Distill, Student Model=Qwen2026.0329.7
No finetuningSurrogate=–, Student Model=Qwen2026.0328.3
Surrogate-TraceStudent Model=Llama, Surrogate=R1, Fine-tuning=true, Victim Model=GPT-5 mini2026.0328.1
Synthesized-TraceSurrogate=R1, Student Model=Llama2026.0325.2
Answer-onlySurrogate=–, Student Model=Qwen2026.0321.6
Answer-onlyStudent Model=Qwen, Surrogate=None, Fine-tuning=true, Victim Model=GPT-5 mini2026.0321.5
Synthesized-TraceStudent Model=Llama, Surrogate=R1, Fine-tuning=true, Victim Model=GPT-5 mini2026.0319.9
Surrogate-TraceStudent Model=Qwen, Surrogate=R1-Distill, Fine-tuning=true, Victim Model=GPT-5 mini2026.0319.7
Surrogate-Trace(R1-Distill)Surrogate=–, Student Model=Qwen2026.0319.7
Answer-onlySurrogate=–, Student Model=Llama2026.0317.2
Synthesized-TraceStudent Model=Llama, Surrogate=R1-Distill, Fine-tuning=true, Victim Model=GPT-5 mini2026.0316.8
No finetuningSurrogate=–, Student Model=Llama2026.0316.6
Synthesized-TraceSurrogate=R1-Distill, Student Model=Llama2026.0316.3
Answer+SummaryStudent Model=Qwen, Surrogate=None, Fine-tuning=true, Victim Model=GPT-5 mini2026.0311.7
Surrogate-TraceStudent Model=Llama, Surrogate=R1-Distill, Fine-tuning=true, Victim Model=GPT-5 mini2026.0311.3
Surrogate-Trace(R1-Distill)Surrogate=–, Student Model=Llama2026.0311.3
Answer-onlyStudent Model=Llama, Surrogate=None, Fine-tuning=true, Victim Model=GPT-5 mini2026.0310
Answer+SummaryStudent Model=Llama, Surrogate=None, Fine-tuning=true, Victim Model=GPT-5 mini2026.038.1