List your company

Language-to-Execution Grounding on LabGuard-Bench stress benchmark (Source-held-out)

76.5Tag F1LabGuard-Hybrid
43.84452.32260.869.278Jun 30, 2026

Evaluation results

MethodLinks
LabGuard-HybridGrounding Module=LabGuard-Hybrid2026.0676.520.4—58.284.168.7
LoRA (Qwen3-8B)Grounding Module=LoRA, Backbone=Qwen3-8B2026.0675.720.6—56.883.467.5
RegexGrounding Module=Regex2026.0645.133.3—47.177.858.7