{ "schema_version": "toolcall-verifier-classifier-diagnostics/v1", "latest_failed_t4_fast_baseline": { "test_macro_f1": 0.3716486037399564, "test_valid_recall": 0.5279547062986554, "test_wrong_tool_semantic_precision": 0.3870967741935484 }, "validation_gate_metrics": { "rows": 11293, "prediction_column": "pred_label", "confidence_column": "confidence", "false_objection_threshold": 0.9, "valid_rows": 3928, "valid_recall": 0.9814154786150713, "valid_false_objection_at_0_90": { "false_objections": 18, "valid_rows": 3928, "rate": 0.004582484725050916 }, "wrong_tool_semantic_precision": 0.9868630573248408, "wrong_tool_semantic_predicted_rows": 2512, "wrong_arguments_semantic_precision": 0.9890520694259012, "wrong_arguments_semantic_predicted_rows": 3745, "valid_to_wrong_arguments_semantic": { "rows": 37, "valid_rows": 3928, "rate": 0.009419551934826884 }, "wrong_tool_to_wrong_arguments_semantic": { "rows": 1, "wrong_tool_rows": 2540, "rate": 0.0003937007874015748 }, "needs_clarification_support": 41 }, "test_gate_metrics": { "rows": 22370, "prediction_column": "pred_label", "confidence_column": "confidence", "false_objection_threshold": 0.9, "valid_rows": 7836, "valid_recall": 0.9823889739663093, "valid_false_objection_at_0_90": { "false_objections": 52, "valid_rows": 7836, "rate": 0.006636038795303726 }, "wrong_tool_semantic_precision": 0.98904052936311, "wrong_tool_semantic_predicted_rows": 4836, "wrong_arguments_semantic_precision": 0.9878325979409012, "wrong_arguments_semantic_predicted_rows": 7479, "valid_to_wrong_arguments_semantic": { "rows": 81, "valid_rows": 7836, "rate": 0.010336906584992343 }, "wrong_tool_to_wrong_arguments_semantic": { "rows": 4, "wrong_tool_rows": 4921, "rate": 0.000812842918106076 }, "needs_clarification_support": 101 }, "validation_deployment_gate_metrics": { "rows": 11293, "prediction_column": "deployment_pred_label", "confidence_column": "deployment_confidence", "false_objection_threshold": null, "valid_rows": 3928, "valid_recall": 0.9989816700610998, "valid_false_objection_deployment": { "false_objections": 4, "valid_rows": 3928, "rate": 0.0010183299389002036 }, "wrong_tool_semantic_precision": 1.0, "wrong_tool_semantic_predicted_rows": 2416, "wrong_arguments_semantic_precision": 0.9988823693769209, "wrong_arguments_semantic_predicted_rows": 3579, "valid_to_wrong_arguments_semantic": { "rows": 4, "valid_rows": 3928, "rate": 0.0010183299389002036 }, "wrong_tool_to_wrong_arguments_semantic": { "rows": 0, "wrong_tool_rows": 2540, "rate": 0.0 }, "needs_clarification_support": 41 }, "test_deployment_gate_metrics": { "rows": 22370, "prediction_column": "deployment_pred_label", "confidence_column": "deployment_confidence", "false_objection_threshold": null, "valid_rows": 7836, "valid_recall": 0.9961715160796325, "valid_false_objection_deployment": { "false_objections": 30, "valid_rows": 7836, "rate": 0.0038284839203675345 }, "wrong_tool_semantic_precision": 0.998502353444587, "wrong_tool_semantic_predicted_rows": 4674, "wrong_arguments_semantic_precision": 0.9967890548652799, "wrong_arguments_semantic_predicted_rows": 7163, "valid_to_wrong_arguments_semantic": { "rows": 23, "valid_rows": 7836, "rate": 0.00293517100561511 }, "wrong_tool_to_wrong_arguments_semantic": { "rows": 0, "wrong_tool_rows": 4921, "rate": 0.0 }, "needs_clarification_support": 101 }, "validation_confusion_pairs": [ { "true_label": "wrong_arguments_semantic", "pred_label": "valid", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 62 }, { "true_label": "wrong_tool_semantic", "pred_label": "valid", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 40 }, { "true_label": "valid", "pred_label": "wrong_arguments_semantic", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 31 }, { "true_label": "valid", "pred_label": "wrong_tool_semantic", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 19 }, { "true_label": "wrong_tool_semantic", "pred_label": "valid", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 18 }, { "true_label": "needs_clarification", "pred_label": "valid", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 9 }, { "true_label": "valid", "pred_label": "wrong_tool_semantic", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 8 }, { "true_label": "wrong_arguments_semantic", "pred_label": "valid", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 6 }, { "true_label": "valid", "pred_label": "wrong_arguments_semantic", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 5 }, { "true_label": "wrong_arguments_semantic", "pred_label": "valid", "source": "forge_fixed_width_numeric", "generator": "unknown", "rows": 5 }, { "true_label": "wrong_arguments_semantic", "pred_label": "valid", "source": "glaiveai/glaive-function-calling-v2", "generator": "unknown", "rows": 4 }, { "true_label": "valid", "pred_label": "tool_not_needed", "source": "agent_training_hf", "generator": "unknown", "rows": 4 }, { "true_label": "tool_not_needed", "pred_label": "wrong_tool_semantic", "source": "agent_training_hf", "generator": "unknown", "rows": 4 }, { "true_label": "needs_clarification", "pred_label": "valid", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 3 }, { "true_label": "needs_clarification", "pred_label": "wrong_arguments_semantic", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 3 }, { "true_label": "valid", "pred_label": "needs_clarification", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 3 }, { "true_label": "wrong_arguments_semantic", "pred_label": "wrong_tool_semantic", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 2 }, { "true_label": "wrong_arguments_semantic", "pred_label": "valid", "source": "agent_training_hf", "generator": "unknown", "rows": 2 }, { "true_label": "valid", "pred_label": "needs_clarification", "source": "glaiveai/glaive-function-calling-v2", "generator": "unknown", "rows": 1 }, { "true_label": "needs_clarification", "pred_label": "valid", "source": "glaiveai/glaive-function-calling-v2", "generator": "unknown", "rows": 1 }, { "true_label": "valid", "pred_label": "needs_clarification", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 1 }, { "true_label": "valid", "pred_label": "wrong_arguments_semantic", "source": "glaiveai/glaive-function-calling-v2", "generator": "unknown", "rows": 1 }, { "true_label": "wrong_tool_semantic", "pred_label": "tool_not_needed", "source": "agent_training_hf", "generator": "unknown", "rows": 1 }, { "true_label": "wrong_tool_semantic", "pred_label": "valid", "source": "glaiveai/glaive-function-calling-v2", "generator": "unknown", "rows": 1 }, { "true_label": "wrong_tool_semantic", "pred_label": "wrong_arguments_semantic", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 1 } ], "test_confusion_pairs": [ { "true_label": "wrong_arguments_semantic", "pred_label": "valid", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 126 }, { "true_label": "wrong_tool_semantic", "pred_label": "valid", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 73 }, { "true_label": "valid", "pred_label": "wrong_arguments_semantic", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 60 }, { "true_label": "wrong_tool_semantic", "pred_label": "valid", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 58 }, { "true_label": "needs_clarification", "pred_label": "valid", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 30 }, { "true_label": "valid", "pred_label": "wrong_tool_semantic", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 29 }, { "true_label": "valid", "pred_label": "wrong_tool_semantic", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 18 }, { "true_label": "valid", "pred_label": "wrong_arguments_semantic", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 16 }, { "true_label": "wrong_arguments_semantic", "pred_label": "valid", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 15 }, { "true_label": "needs_clarification", "pred_label": "valid", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 11 }, { "true_label": "needs_clarification", "pred_label": "valid", "source": "glaiveai/glaive-function-calling-v2", "generator": "unknown", "rows": 9 }, { "true_label": "needs_clarification", "pred_label": "wrong_arguments_semantic", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 5 }, { "true_label": "wrong_arguments_semantic", "pred_label": "valid", "source": "glaiveai/glaive-function-calling-v2", "generator": "unknown", "rows": 5 }, { "true_label": "valid", "pred_label": "wrong_arguments_semantic", "source": "glaiveai/glaive-function-calling-v2", "generator": "unknown", "rows": 4 }, { "true_label": "wrong_arguments_semantic", "pred_label": "wrong_tool_semantic", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 2 }, { "true_label": "valid", "pred_label": "tool_not_needed", "source": "agent_training_hf", "generator": "unknown", "rows": 2 }, { "true_label": "valid", "pred_label": "needs_clarification", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 2 }, { "true_label": "tool_not_needed", "pred_label": "wrong_tool_semantic", "source": "agent_training_hf", "generator": "unknown", "rows": 2 }, { "true_label": "valid", "pred_label": "needs_clarification", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 2 }, { "true_label": "valid", "pred_label": "needs_clarification", "source": "glaiveai/glaive-function-calling-v2", "generator": "unknown", "rows": 2 }, { "true_label": "wrong_tool_semantic", "pred_label": "wrong_arguments_semantic", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 2 }, { "true_label": "wrong_tool_semantic", "pred_label": "valid", "source": "glaiveai/glaive-function-calling-v2", "generator": "unknown", "rows": 2 }, { "true_label": "wrong_arguments_semantic", "pred_label": "needs_clarification", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 2 }, { "true_label": "wrong_tool_semantic", "pred_label": "wrong_arguments_semantic", "source": "Salesforce/xlam-function-calling-60k", "generator": "unknown", "rows": 2 }, { "true_label": "tool_not_needed", "pred_label": "valid", "source": "agent_training_hf", "generator": "unknown", "rows": 1 }, { "true_label": "needs_clarification", "pred_label": "wrong_tool_semantic", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 1 }, { "true_label": "needs_clarification", "pred_label": "wrong_arguments_semantic", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 1 }, { "true_label": "valid", "pred_label": "tool_not_needed", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 1 }, { "true_label": "valid", "pred_label": "wrong_tool_semantic", "source": "glaiveai/glaive-function-calling-v2", "generator": "unknown", "rows": 1 }, { "true_label": "valid", "pred_label": "wrong_arguments_semantic", "source": "forge_trace", "generator": "unknown", "rows": 1 }, { "true_label": "wrong_arguments_semantic", "pred_label": "valid", "source": "forge_fixed_width_numeric", "generator": "unknown", "rows": 1 }, { "true_label": "wrong_arguments_semantic", "pred_label": "tool_not_needed", "source": "agent_training_hf", "generator": "unknown", "rows": 1 }, { "true_label": "wrong_arguments_semantic", "pred_label": "tool_not_needed", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 1 }, { "true_label": "wrong_arguments_semantic", "pred_label": "valid", "source": "forge_error_recovery_numeric", "generator": "unknown", "rows": 1 }, { "true_label": "wrong_arguments_semantic", "pred_label": "valid", "source": "agent_training_hf", "generator": "unknown", "rows": 1 }, { "true_label": "wrong_tool_semantic", "pred_label": "tool_not_needed", "source": "Team-ACE/ToolACE", "generator": "unknown", "rows": 1 } ] }