List your company

Intent Grasping on IntentGrasp All Set (downsample)

61.68F1 ScoreGemini 3.1-Pro
37.65643.89350.1356.367May 7, 2026

Evaluation results

MethodLinks
Gemini 3.1-Prosample=sample-432026.0561.68——
Gemini 3.1-Prosample=sample-442026.0560.41——
Claude Opus-4.7sample=sample-442026.0560.07——
Claude Opus-4.7sample=sample-432026.0558.41——
Gemini 3-Flashsample=sample-442026.0557.08——
Gemini 3.1-Prosample=sample-422026.0556.96——
Gemini 3-Flashsample=sample-432026.0555.55——
Gemini 3-Flashsample=sample-422026.0553.57——
Claude Opus-4.7sample=sample-422026.0553.44——
Claude Sonnet-4.6sample=sample-432026.0553.2——
GPT 5.4sample=sample-442026.0553.17——
Gemini 3.1-Flash-Litesample=sample-432026.0552.45——
Gemini 3.1-Flash-Litesample=sample-442026.0552.14——
GPT 5.4sample=sample-432026.0551.18——
Claude Sonnet-4.6sample=sample-442026.0550.35——
Claude Haiku-4.5sample=sample-442026.0549.25——
Gemini 3.1-Flash-Litesample=sample-422026.0548.38——
Claude Haiku-4.5sample=sample-432026.0548.35——
GPT 5.4sample=sample-422026.0547.71——
Claude Sonnet-4.6sample=sample-422026.0547.59——
GPT 5.4-minisample=sample-442026.0546.13——
GPT 5.4-minisample=sample-432026.0545.32——
GPT 5.4-nanosample=sample-442026.0543.7——
Claude Haiku-4.5sample=sample-422026.0542.12——
GPT 5.4-minisample=sample-422026.0540.68——
GPT 5.4-nanosample=sample-422026.0538.77——
GPT 5.4-nanosample=sample-432026.0538.58——
Claude Haiku-4.5aggregation=Average across samples2026.05—46.573.17
Claude Opus-4.7aggregation=Average across samples2026.05—57.312.82
Claude Sonnet-4.6aggregation=Average across samples2026.05—50.382.29
Gemini 3-Flashaggregation=Average across samples2026.05—55.41.44
Gemini 3.1-Flash-Liteaggregation=Average across samples2026.05—50.991.85
Gemini 3.1-Proaggregation=Average across samples2026.05—59.681.99
GPT 5.4aggregation=Average across samples2026.05—50.692.26
GPT 5.4-miniaggregation=Average across samples2026.05—44.042.4
GPT 5.4-nanoaggregation=Average across samples2026.05—40.352.37