List your company
ResearchFormerly WizWand

Logical Reasoning on AR-LSAT (test)

91.66AccuracyMIPROv2
17.487236.74365675.2564Dec 3, 2025Dec 13, 2025Dec 23, 2025Jan 2, 2026Jan 12, 2026Jan 22, 2026Feb 1, 2026
Updated 8mo ago

Evaluation results

MethodLinks
MIPROv2Backbone model=DeepSeek-V3.1, Optimizer model=GPT-4.12026.0291.66—
ETGPOBackbone model=DeepSeek-V3.1, Optimizer model=GPT-4.12026.0291.44—
GEPABackbone model=DeepSeek-V3.1, Optimizer model=GPT-4.12026.0290.92—
Chain of ThoughtBackbone model=DeepSeek-V3.1, Optimizer model=GPT-4.12026.0290.65—
HBLRModel=DeepSeek-R1, Prompting Strategy=HBLR2025.1286.47—
CoTModel=DeepSeek-R1, Prompting Strategy=CoT2025.1276.91—
SymbCoTModel=DeepSeek-R1, Prompting Strategy=SymbCoT2025.1270.87—
LogicModel=DeepSeek-R1, Prompting Strategy=Logic2025.1262.1—
DirectModel=DeepSeek-R1, Prompting Strategy=Direct2025.1260.4—
HBLRModel=DeepSeek-V3, Prompting Strategy=HBLR2025.1246.69—
HBLRModel=GPT-4, Prompting Strategy=HBLR2025.1244.67—
SymbCoTModel=DeepSeek-V3, Prompting Strategy=SymbCoT2025.1244.52—
LogicModel=GPT-4, Prompting Strategy=Logic2025.1243.04—
CoTModel=DeepSeek-V3, Prompting Strategy=CoT2025.1242.5—
LogicModel=DeepSeek-V3, Prompting Strategy=Logic2025.1240.41—
SymbCoTModel=GPT-4, Prompting Strategy=SymbCoT2025.1237.19—
DirectModel=DeepSeek-V3, Prompting Strategy=Direct2025.1236.21—
CoTModel=GPT-4, Prompting Strategy=CoT2025.1235.06—
DirectModel=GPT-4, Prompting Strategy=Direct2025.1234.43—
HBLRModel=GPT-3.5-Turbo, Prompting Strategy=HBLR2025.1227.22—
LogicModel=GPT-3.5-Turbo, Prompting Strategy=Logic2025.1225.15—
SymbCoTModel=GPT-3.5-Turbo, Prompting Strategy=SymbCoT2025.1221.59—
CoTModel=GPT-3.5-Turbo, Prompting Strategy=CoT2025.1221.31—
DirectModel=GPT-3.5-Turbo, Prompting Strategy=Direct2025.1220.34—
ETGPOBackbone=DeepSeek-V3.1, Optimizer=GPT-4.12026.02—1,378
GEPABackbone=DeepSeek-V3.1, Optimizer=GPT-4.12026.02—5,523
MIPROv2Backbone=DeepSeek-V3.1, Optimizer=GPT-4.12026.02—10,735