List your company

Deductive logical reasoning on FOLIO 203 (dev)

6.4Exclusion RateICL
3.85821.016538.17555.3335Jan 14, 2026

Evaluation results

MethodLinks
ICLModel=Qwen3-4B-Instruct-2507, Inference Type=In-context learning (5-shots)2026.016.43.45
ICLModel=Qwen2.5-3B-Instruct, Inference Type=In-context learning (5-shots)2026.0127.5919.21
ICLModel=Phi-4-mini-Instruct, Inference Type=In-context learning (5-shots)2026.0145.3228.08
IncrementModel=Qwen2.5-3B-Instruct, Inference Type=Incremental inference2026.0155.1730.54
SFT+Model=Qwen2.5-3B-Instruct, Inference Type=SFT on symbolic ProofWriter and FOLIO combined2026.0157.6330.05
ICLModel=Gemma-3-4B-Instruct, Inference Type=In-context learning (5-shots)2026.0158.1333
IncrementModel=Qwen3-4B-Instruct-2507, Inference Type=Incremental inference2026.0158.1338.42
SFT+Model=Qwen3-4B-Instruct-2507, Inference Type=SFT on symbolic ProofWriter and FOLIO combined2026.0160.140.39
SFT+Model=Phi-4-mini-Instruct, Inference Type=SFT on symbolic ProofWriter and FOLIO combined2026.0161.0838.92
IncrementModel=Phi-4-mini-Instruct, Inference Type=Incremental inference2026.0162.0738.92
SFT+Model=Gemma-3-4B-Instruct, Inference Type=SFT on symbolic ProofWriter and FOLIO combined2026.016636.95
IncrementModel=Gemma-3-4B-Instruct, Inference Type=Incremental inference2026.0169.9540.89