List your company

STEM Question Answering on TheoremQA

76.2CorrectnessMOCHA
-2.89907217.63626438.171658.706936May 19, 2026May 22, 2026May 26, 2026May 30, 2026Jun 2, 2026Jun 6, 2026Jun 10, 2026

Evaluation results

MethodLinks
MOCHAEvaluation Backbone=Claude Haiku 4.5, Rollout Budget=1000, Number of Seeds=52026.0576.2—
ProTeGiEvaluation Backbone=Claude Haiku 4.5, Rollout Budget=1000, Number of Seeds=52026.0569—
TextGradEvaluation Backbone=Claude Haiku 4.5, Rollout Budget=1000, Number of Seeds=52026.0567.2—
GEPAEvaluation Backbone=Claude Haiku 4.5, Rollout Budget=1000, Number of Seeds=52026.0565.6—
Seed SkillEvaluation Backbone=Claude Haiku 4.5, Rollout Budget=1000, Number of Seeds=52026.0553.4—
Full TextBackbone Model=Qwen3-8B2026.060.6881935
SKIM (Adaptive)Backbone Model=Qwen3-8B2026.060.6787727
LLMLingua-2 (L)Backbone Model=Qwen3-8B2026.060.6613710
SKIM (Fix-512)Backbone Model=Qwen3-8B2026.060.6479509
LLMLingua-2 (M)Backbone Model=Qwen3-8B2026.060.6292535
SKIM (Fix-256)Backbone Model=Qwen3-8B2026.060.6024256
LLMLingua-2 (S)Backbone Model=Qwen3-8B2026.060.5649281
NaiveBackbone Model=Qwen3-8B2026.060.52070
Full TextBackbone Model=Phi-42026.060.5114921
LLMLingua-2 (L)Backbone Model=Phi-42026.060.4873693
SKIM (Adaptive)Backbone Model=Phi-42026.060.4873682
SKIM (Fix-512)Backbone Model=Phi-42026.060.4833509
LLMLingua-2 (S)Backbone Model=Phi-42026.060.4739272
LLMLingua-2 (M)Backbone Model=Phi-42026.060.4565521
NaiveBackbone Model=Phi-42026.060.45380
SKIM (Fix-256)Backbone Model=Phi-42026.060.4378256
ICAEBackbone Model=Qwen3-8B2026.060.2838256
ICAEBackbone Model=Phi-42026.060.1861256
500xCompressorBackbone Model=Phi-42026.060.1513256
500xCompressorBackbone Model=Qwen3-8B2026.060.1432256