List your company
ResearchFormerly WizWand

Reasoning on MMLU-Pro (Score %)

78MMLU-Pro Reasoning ScoreDR-Rubric-8B (BS-3)
29.6442.19554.7567.305Apr 17, 2026Apr 24, 2026May 1, 2026May 9, 2026May 16, 2026May 23, 2026May 31, 2026
Updated 5mo ago

Evaluation results

MethodLinks
DR-Rubric-8B (BS-3)Training=SFT+RL, 3K2026.0578
DR-Rubric-8B (Gemini)Training=SFT+RL, 1K2026.0575
DR-Rubric-8B (GPT-5)Training=SFT+RL, 1K2026.0574.8
Qwen2.5-7B2026.0574.3
Qwen3-8B-SFTTraining=SFT, 1K2026.0573.8
Qwen3-8B2026.0573.3
WebExplorer-8BTraining=SFT+RL, 25K2026.0571.5
DR-Tulu-SFT-8BTraining=SFT, 16K2026.0571
Pair+Backbone=olmo-3.1-32b-it-sft2026.0469.4
MPO+Backbone=olmo-3.1-32b-it-sft2026.0469.3
Soft+Backbone=olmo-3.1-32b-it-sft2026.0469.1
DR-Tulu-RL-8BTraining=SFT+RL, 25K2026.0569
Search-R1-7BTraining=RL, 90K2026.0568
DPOBackbone=olmo-3.1-32b-it-sft2026.0467.9
Mag+Backbone=olmo-3.1-32b-it-sft2026.0467.3
RFTBackbone=olmo-3.1-32b-it-sft2026.0458.2
BaseBackbone=olmo-3.1-32b-it-sft2026.0457
Pair+Backbone=olmo-3-7b-it-sft2026.0455.2
TeacherMethod ID=C0, Distillation Strategy=Teacher, Training Stage=Before SFT+DPO2026.0555.2
Mag+Backbone=olmo-3-7b-it-sft2026.0455
RSAMethod ID=C3, Distillation Strategy=RSA, Training Stage=Before SFT+DPO2026.0555
HEEDMethod ID=C4, Distillation Strategy=HEED, Training Stage=Before SFT+DPO2026.0555
MPO+Backbone=olmo-3-7b-it-sft2026.0454.8
HSAMethod ID=C2, Distillation Strategy=HSA, Training Stage=Before SFT+DPO2026.0554.4
KDMethod ID=C1, Distillation Strategy=KD, Training Stage=Before SFT+DPO2026.0553.9
Soft+Backbone=olmo-3-7b-it-sft2026.0453.6
DPOBackbone=olmo-3-7b-it-sft2026.0451.6
RFTBackbone=olmo-3-7b-it-sft2026.0446.3
BaseBackbone=olmo-3-7b-it-sft2026.0445.6
Soft+Backbone=gemma-3-4b-sft2026.0437.5
DPOBackbone=gemma-3-4b-sft2026.0436.7
MPO+Backbone=gemma-3-4b-sft2026.0436.2
Mag+Backbone=gemma-3-4b-sft2026.0434.7
Pair+Backbone=gemma-3-4b-sft2026.0433.5
RFTBackbone=gemma-3-4b-sft2026.0432.8
BaseBackbone=gemma-3-4b-sft2026.0431.5