List your company

Vignette Completeness on CIRL-Vig

45.8Vignette CompletenessOpenThinker3-7B
-1.83210.53422.935.266Apr 21, 2026

Evaluation results

MethodLinks
OpenThinker3-7BCondition=Zero-shot2026.0445.8
Gemma-3-12BCondition=SFT2026.0445.4
Gemma-3-12BCondition=Zero-shot2026.0443.6
Phi-4Condition=Zero-shot2026.0443
Qwen3.5-9BCondition=SFT2026.0441.6
Qwen3.5-2BCondition=Zero-shot2026.0441.6
Qwen3.5-4BCondition=Zero-shot2026.0441.6
Qwen3.5-2BCondition=SFT2026.0440.8
Qwen3.5-9BCondition=Zero-shot2026.0438.7
OpenThinker3-7BCondition=SFT2026.0434.1
Phi-4Condition=SFT2026.0427.8
ContextReasoner-7BCondition=Zero-shot2026.0413
Qwen3.5-9BCondition=GRPO2026.040
GPT-OSS-20BCondition=Zero-shot2026.040
GPT-OSS-20BCondition=SFT2026.040