List your company

Natural Language Inference on ContractNLI re-annotated (test)

83AccuracyClaude
38.2849.8961.573.11Jun 15, 2026

Evaluation results

MethodLinks
ClaudeClassification Approach=LLM-Formal Reasoning2026.0683
ClaudeClassification Approach=Solver-Formal Reasoning2026.0674.5
QwenClassification Approach=LLM-Formal Reasoning2026.0669.1
DeepseekClassification Approach=Simple2026.0668.4
GPTClassification Approach=Simple2026.0666
GPTClassification Approach=LLM-Formal Reasoning2026.0665.3
QwenClassification Approach=Simple2026.0665.1
DeepseekClassification Approach=LLM-Formal Reasoning2026.0664.2
ClaudeClassification Approach=Simple2026.0663.1
DeepseekClassification Approach=Structured2026.0662.2
QwenClassification Approach=Structured2026.0661.5
GPTClassification Approach=Solver-Formal Reasoning2026.0660.3
GPTClassification Approach=Structured2026.0658.8
LlamaClassification Approach=LLM-Formal Reasoning2026.0658.6
LlamaClassification Approach=Simple2026.0655.9
ClaudeClassification Approach=Structured2026.0653.2
QwenClassification Approach=Solver-Formal Reasoning2026.0648.3
DeepseekClassification Approach=Solver-Formal Reasoning2026.0645.3
LlamaClassification Approach=Solver-Formal Reasoning2026.0642.1
LlamaClassification Approach=Structured2026.0640