List your company

Customer Support BPM on Customer Support BPM Dataset (test)

79.15Tool Accuracypolicy–critic pipeline
33.18245.11657.0568.984Apr 26, 2026

Evaluation results

MethodLinks
policy–critic pipelineEvaluation Protocol=domain-specific SFT2026.0479.1565.48
Gemini 3 ProEvaluation Protocol=zero-shot prompting2026.0447.7832.93
Qwen3-235B-A22BEvaluation Protocol=zero-shot prompting2026.0444.3326.18
GPT-5.2Evaluation Protocol=zero-shot prompting2026.0443.0127.02
DeepSeek V3.1Evaluation Protocol=zero-shot prompting2026.0442.427.83
Qwen2.5-Instruct-7B (baseline)Evaluation Protocol=zero-shot prompting2026.0434.9510.27