eval-harness

metric

model_triage_cost

better
lower
kind
quality: measured against a reference or by a judge
decides
no goal
weighed by
no goal
recorded on
orrery_triage_v1_labelled

01what it measures, and how

triage_cost of the model's own decision, before the fleet's gates changed it.

A project explains its own metric at length with METRIC_DESCRIPTIONS = {"model_triage_cost": {"summary": "…", "details": "…"}}.

02on orrery_triage_v1_labelled

gold/orrery_triage_v1_labelledtriage_stub_escalate_v1: 0.477triage_stub_escalate_v1 · rescored 3d3a6afd: 0.477triage_stub_escalate_v1 · rescored 5f2c057c: 0.477triage_stub_escalate_v1 · rescored 77adc97b: 0.477triage_stub_escalate_v1 · rescored cd35dc1f: 0.477triage_gpt_61_sol_v1 · rescored 3d3a6afd, repeats ×3: 0.523triage_gpt_61_sol_v1 · rescored 5f2c057c, repeats ×3: 0.523triage_gpt_61_sol_v1 · rescored 77adc97b, repeats ×3: 0.523triage_gpt_61_sol_v1 · rescored cd35dc1f, repeats ×3: 0.523triage_gpt_61_sol_v1 · rescored 3d3a6afd: 0.554triage_gpt_61_sol_v1 · rescored 5f2c057c: 0.554triage_gpt_61_sol_v1 · rescored 77adc97b: 0.554triage_gpt_61_sol_v1 · rescored cd35dc1f: 0.554triage_gpt_6_luna_v1 · rescored 5f2c057c: 0.615triage_gpt_6_luna_v1 · rescored 77adc97b: 0.615triage_gpt_6_luna_v1 · rescored cd35dc1f: 0.615triage_gpt_6_luna_v1 · rescored 5f2c057c, repeats ×3: 0.682triage_gpt_6_luna_v1 · rescored 77adc97b, repeats ×3: 0.682triage_gpt_6_luna_v1 · rescored cd35dc1f, repeats ×3: 0.682triage_local_dgx_phi4_14b_v1: 0.785triage_local_mac_phi4_14b_v1: 0.785triage_local_dgx_phi4_14b_v1 · rescored 3d3a6afd: 0.785triage_local_mac_phi4_14b_v1 · rescored 3d3a6afd: 0.785triage_local_dgx_phi4_14b_v1 · rescored 5f2c057c: 0.785triage_local_mac_phi4_14b_v1 · rescored 5f2c057c: 0.785triage_local_dgx_phi4_14b_v1 · rescored 77adc97b: 0.785triage_local_mac_phi4_14b_v1 · rescored 77adc97b: 0.785triage_local_dgx_phi4_14b_v1 · rescored cd35dc1f: 0.785triage_local_mac_phi4_14b_v1 · rescored cd35dc1f: 0.785triage_gpt_6_luna_v2: 0.831triage_gpt_6_luna_v2 · rescored 3d3a6afd: 0.831triage_gpt_6_luna_v2 · rescored 5f2c057c: 0.831triage_local_x3_phi4_14b_v1: 0.862triage_local_x3_phi4_14b_v1 · rescored 3d3a6afd: 0.862triage_local_x3_phi4_14b_v1 · rescored 5f2c057c: 0.862triage_local_x3_phi4_14b_v1 · rescored 77adc97b: 0.862triage_local_x3_phi4_14b_v1 · rescored cd35dc1f: 0.862triage_haiku_55_v1 · rescored 3d3a6afd: 0.938triage_haiku_55_v1 · rescored 5f2c057c: 0.938triage_haiku_55_v1 · rescored 77adc97b: 0.938triage_haiku_55_v1 · rescored cd35dc1f: 0.938triage_local_x3_gpt_oss_20b_v1: 0.969triage_haiku_55_v1 · rescored 3d3a6afd, repeats ×3: 0.969triage_kimi_k31_pro_v1 · rescored 5f2c057c: 0.969triage_haiku_55_v1 · rescored 5f2c057c, repeats ×3: 0.969triage_kimi_k31_pro_v1 · rescored 77adc97b: 0.969triage_haiku_55_v1 · rescored 77adc97b, repeats ×3: 0.969triage_kimi_k31_pro_v1 · rescored cd35dc1f: 0.969triage_local_x3_gpt_oss_20b_v1 · rescored cd35dc1f: 0.969triage_haiku_55_v1 · rescored cd35dc1f, repeats ×3: 0.969triage_local_x3_nemotron_35_bolt_e4b_v1: 1.015triage_local_x3_nemotron_35_bolt_e4b_v1 · repeats: 1.015triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c: 1.015triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c, repeats: 1.015triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b: 1.015triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b, repeats: 1.015triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f: 1.015triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f, repeats: 1.015triage_local_x3_nemotron_35_lightning_30b_v1: 1.046triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 5f2c057c: 1.046triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 77adc97b: 1.046triage_local_x3_nemotron_35_lightning_30b_v1 · rescored cd35dc1f: 1.046triage_local_mac_llama31_8b_v1: 1.062triage_local_mac_llama31_8b_v1 · rescored 3d3a6afd: 1.062triage_local_x3_llama31_8b_v1 · rescored 3d3a6afd: 1.062triage_local_mac_llama31_8b_v1 · rescored 5f2c057c: 1.062triage_local_x3_llama31_8b_v1 · rescored 5f2c057c: 1.062triage_local_mac_llama31_8b_v1 · rescored 77adc97b: 1.062triage_local_x3_llama31_8b_v1 · rescored 77adc97b: 1.062triage_local_mac_llama31_8b_v1 · rescored cd35dc1f: 1.062triage_local_x3_llama31_8b_v1 · rescored cd35dc1f: 1.062triage_local_dgx_llama31_8b_v1: 1.092triage_local_dgx_llama31_8b_v1 · rescored 3d3a6afd: 1.092triage_local_dgx_llama31_8b_v1 · rescored 5f2c057c: 1.092triage_local_dgx_llama31_8b_v1 · rescored 77adc97b: 1.092triage_local_dgx_llama31_8b_v1 · rescored cd35dc1f: 1.092triage_local_dgx_nemotron_35_lightning_30b_v1: 1.138triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1: 1.138triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 3d3a6afd: 1.138triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 5f2c057c: 1.138triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 5f2c057c: 1.138triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 77adc97b: 1.138triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 77adc97b: 1.138triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored cd35dc1f: 1.138triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored cd35dc1f: 1.138triage_gemini_4_argon_v1 · rescored 3d3a6afd: 1.185triage_gemini_4_argon_v1 · rescored 5f2c057c: 1.185triage_gemini_4_argon_v1 · rescored 77adc97b: 1.185triage_gemini_4_argon_v1 · rescored cd35dc1f: 1.185triage_local_dgx_nemotron_35_bolt_e4b_v1: 1.246triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c: 1.246triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 77adc97b: 1.246triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f: 1.246triage_local_dgx_gpt_oss_20b_v1: 1.277triage_local_dgx_gpt_oss_20b_v1 · rescored cd35dc1f: 1.277triage_local_mac_nemotron_35_bolt_e4b_v1: 1.292triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c: 1.292triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 77adc97b: 1.292triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f: 1.292triage_local_mac_gpt_oss_20b_v1: 1.431triage_local_mac_gpt_oss_20b_v1 · rescored cd35dc1f: 1.431triage_local_mac_nemotron_35_lightning_30b_v1: 1.446triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 5f2c057c: 1.446triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 77adc97b: 1.446triage_local_mac_nemotron_35_lightning_30b_v1 · rescored cd35dc1f: 1.446triage_local_dgx_nemotron_35_super_120b_v2: 1.462triage_local_mac_nemotron_35_lightning_30b_v2: 1.462triage_local_x3_nemotron_35_lightning_30b_v2: 1.462triage_local_dgx_nemotron_35_super_120b_v2 · rescored 3d3a6afd: 1.462triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd: 1.462triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd: 1.462triage_local_dgx_nemotron_35_super_120b_v2 · rescored 5f2c057c: 1.462triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 5f2c057c: 1.462triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 5f2c057c: 1.462triage_kimi_k31_pro_v2: 1.477triage_kimi_k31_pro_v2 · rescored 3d3a6afd: 1.477triage_kimi_k31_pro_v2 · rescored 5f2c057c: 1.477triage_sonnet_55_v1 · rescored 3d3a6afd: 1.492triage_sonnet_55_v1 · rescored 5f2c057c: 1.492triage_sonnet_55_v1 · rescored 77adc97b: 1.492triage_sonnet_55_v1 · rescored cd35dc1f: 1.492triage_hosted_qwen3_30b_a3b_v1: 1.508triage_local_dgx_nemotron_35_super_120b_v1: 1.508triage_hosted_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 1.508triage_hosted_qwen3_30b_a3b_v1 · rescored 5f2c057c: 1.508triage_local_dgx_nemotron_35_super_120b_v1 · rescored 5f2c057c: 1.508triage_hosted_qwen3_30b_a3b_v1 · rescored 77adc97b: 1.508triage_local_dgx_nemotron_35_super_120b_v1 · rescored 77adc97b: 1.508triage_hosted_qwen3_30b_a3b_v1 · rescored cd35dc1f: 1.508triage_local_dgx_nemotron_35_super_120b_v1 · rescored cd35dc1f: 1.508triage_local_dgx_gpt_oss_120b_v1: 1.523triage_local_mac_gpt_oss_20b_v2: 1.523triage_local_mac_gpt_oss_20b_v2 · rescored 3d3a6afd: 1.523triage_local_dgx_gpt_oss_120b_v1 · rescored 5f2c057c: 1.523triage_local_mac_gpt_oss_20b_v2 · rescored 5f2c057c: 1.523triage_local_dgx_gpt_oss_120b_v1 · rescored 77adc97b: 1.523triage_local_dgx_gpt_oss_120b_v1 · rescored cd35dc1f: 1.523triage_local_dgx_nemotron_35_lightning_30b_v2: 1.538triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd: 1.538triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 5f2c057c: 1.538triage_local_dgx_gpt_oss_20b_v2: 1.554triage_local_mac_nemotron_35_bolt_e4b_v2: 1.554triage_baseline_pro_v1 · rescored 3d3a6afd: 1.554triage_local_dgx_gpt_oss_20b_v2 · rescored 3d3a6afd: 1.554triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd: 1.554triage_baseline_pro_v1 · rescored 5f2c057c: 1.554triage_local_dgx_gpt_oss_20b_v2 · rescored 5f2c057c: 1.554triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c: 1.554triage_baseline_pro_v1 · rescored 77adc97b: 1.554triage_baseline_pro_v1 · rescored cd35dc1f: 1.554triage_local_dgx_llama4_scout_v1: 1.569triage_local_dgx_llama4_scout_v1 · rescored 3d3a6afd: 1.569triage_local_dgx_llama4_scout_v1 · rescored 5f2c057c: 1.569triage_local_dgx_llama4_scout_v1 · rescored 77adc97b: 1.569triage_local_dgx_llama4_scout_v1 · rescored cd35dc1f: 1.569triage_local_dgx_gemma4_12b_noreason_v1: 1.6triage_local_dgx_gemma4_12b_noreason_v1 · rescored 3d3a6afd: 1.6triage_local_dgx_gemma4_12b_noreason_v1 · rescored 5f2c057c: 1.6triage_local_dgx_gemma4_12b_noreason_v1 · rescored 77adc97b: 1.6triage_local_dgx_gemma4_12b_noreason_v1 · rescored cd35dc1f: 1.6triage_local_dgx_gemma4_31b_v1: 1.631triage_local_dgx_gemma4_31b_v1 · rescored 5f2c057c: 1.631triage_local_dgx_gemma4_31b_v1 · rescored 77adc97b: 1.631triage_local_dgx_gemma4_31b_v1 · rescored cd35dc1f: 1.631triage_local_mac_gemma4_31b_v1: 1.646triage_hosted_gpt_oss_120b_v1 · rescored 5f2c057c: 1.646triage_local_mac_gemma4_31b_v1 · rescored 5f2c057c: 1.646triage_hosted_gpt_oss_120b_v1 · rescored 77adc97b: 1.646triage_local_mac_gemma4_31b_v1 · rescored 77adc97b: 1.646triage_hosted_gpt_oss_120b_v1 · rescored cd35dc1f: 1.646triage_local_mac_gemma4_31b_v1 · rescored cd35dc1f: 1.646triage_local_dgx_mistral_medium_35_128b_v1: 1.662triage_local_mac_gemma4_12b_noreason_v1: 1.662triage_hosted_mistral_medium_35_v1 · rescored 3d3a6afd: 1.662triage_local_dgx_mistral_medium_35_128b_v1 · rescored 3d3a6afd: 1.662triage_local_mac_gemma4_12b_noreason_v1 · rescored 3d3a6afd: 1.662triage_hosted_mistral_medium_35_v1 · rescored 5f2c057c: 1.662triage_local_dgx_mistral_medium_35_128b_v1 · rescored 5f2c057c: 1.662triage_local_mac_gemma4_12b_noreason_v1 · rescored 5f2c057c: 1.662triage_local_x3_gpt_oss_20b_v1 · rescored 5f2c057c: 1.662triage_hosted_mistral_medium_35_v1 · rescored 77adc97b: 1.662triage_local_dgx_mistral_medium_35_128b_v1 · rescored 77adc97b: 1.662triage_local_mac_gemma4_12b_noreason_v1 · rescored 77adc97b: 1.662triage_local_x3_gpt_oss_20b_v1 · rescored 77adc97b: 1.662triage_hosted_mistral_medium_35_v1 · rescored cd35dc1f: 1.662triage_local_dgx_mistral_medium_35_128b_v1 · rescored cd35dc1f: 1.662triage_local_mac_gemma4_12b_noreason_v1 · rescored cd35dc1f: 1.662triage_hosted_qwen35_122b_v1 · rescored 5f2c057c: 1.692triage_hosted_qwen35_122b_v1 · rescored 77adc97b: 1.692triage_hosted_qwen35_122b_v1 · rescored cd35dc1f: 1.692triage_local_dgx_nemotron_35_bolt_e4b_v2: 1.708triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd: 1.708triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c: 1.708triage_local_dgx_gpt_oss_120b_v2: 1.723triage_local_x3_gemma4_31b_v1: 1.723triage_local_x3_nemotron_35_bolt_e4b_v2: 1.723triage_local_dgx_gpt_oss_120b_v2 · rescored 3d3a6afd: 1.723triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd: 1.723triage_local_dgx_gpt_oss_120b_v2 · rescored 5f2c057c: 1.723triage_local_x3_gemma4_31b_v1 · rescored 5f2c057c: 1.723triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c: 1.723triage_local_x3_gemma4_31b_v1 · rescored 77adc97b: 1.723triage_local_x3_gemma4_31b_v1 · rescored cd35dc1f: 1.723triage_hosted_gpt_oss_120b_v2: 1.738triage_hosted_gpt_oss_120b_v2 · rescored 3d3a6afd: 1.738triage_hosted_gpt_oss_120b_v2 · rescored 5f2c057c: 1.738triage_baseline_pro_v1 · rescored 3d3a6afd, repeats ×3: 1.764triage_baseline_pro_v1 · rescored 5f2c057c, repeats ×3: 1.764triage_baseline_pro_v1 · rescored 77adc97b, repeats ×3: 1.764triage_baseline_pro_v1 · rescored cd35dc1f, repeats ×3: 1.764triage_local_x3_gpt_oss_20b_v2: 1.785triage_local_x3_gpt_oss_20b_v2 · rescored 3d3a6afd: 1.785triage_local_x3_gpt_oss_20b_v2 · rescored 5f2c057c: 1.785triage_local_dgx_qwen35_122b_v1: 1.8triage_local_dgx_qwen35_122b_v1 · rescored 5f2c057c: 1.8triage_local_dgx_qwen35_122b_v1 · rescored 77adc97b: 1.8triage_local_dgx_qwen35_122b_v1 · rescored cd35dc1f: 1.8triage_local_dgx_gpt_oss_20b_v1 · rescored 5f2c057c: 1.815triage_local_dgx_gpt_oss_20b_v1 · rescored 77adc97b: 1.815triage_local_mac_qwen3_30b_a3b_v1: 1.862triage_local_x3_qwen3_30b_a3b_v1: 1.862triage_local_mac_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 1.862triage_local_x3_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 1.862triage_local_mac_qwen3_30b_a3b_v1 · rescored 5f2c057c: 1.862triage_local_x3_qwen3_30b_a3b_v1 · rescored 5f2c057c: 1.862triage_local_mac_qwen3_30b_a3b_v1 · rescored 77adc97b: 1.862triage_local_x3_qwen3_30b_a3b_v1 · rescored 77adc97b: 1.862triage_local_mac_qwen3_30b_a3b_v1 · rescored cd35dc1f: 1.862triage_local_x3_qwen3_30b_a3b_v1 · rescored cd35dc1f: 1.862triage_mistral_large_4_v2: 1.908triage_mistral_large_4_v2 · rescored 3d3a6afd: 1.908triage_mistral_large_4_v2 · rescored 5f2c057c: 1.908triage_mistral_large_4_v1 · rescored 5f2c057c: 1.923triage_mistral_large_4_v1 · rescored 77adc97b: 1.923triage_mistral_large_4_v1 · rescored cd35dc1f: 1.923triage_local_mac_gpt_oss_20b_v1 · rescored 5f2c057c: 1.969triage_local_mac_gpt_oss_20b_v1 · rescored 77adc97b: 1.969triage_hosted_qwen35_122b_v2: 1.985triage_hosted_qwen35_122b_v2 · rescored 3d3a6afd: 1.985triage_hosted_qwen35_122b_v2 · rescored 5f2c057c: 1.985triage_gemini_35_flash_lite_v1 · rescored 3d3a6afd: 2.031triage_gemini_35_flash_lite_v1 · rescored 5f2c057c: 2.031triage_gemini_35_flash_lite_v1 · rescored 77adc97b: 2.031triage_gemini_35_flash_lite_v1 · rescored cd35dc1f: 2.031triage_local_dgx_qwen3_30b_a3b_v1: 2.108triage_local_dgx_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 2.108triage_local_dgx_qwen3_30b_a3b_v1 · rescored 5f2c057c: 2.108triage_local_dgx_qwen3_30b_a3b_v1 · rescored 77adc97b: 2.108triage_local_dgx_qwen3_30b_a3b_v1 · rescored cd35dc1f: 2.108triage_hosted_gemma4_31b_v1 · rescored 5f2c057c: 2.154triage_hosted_gemma4_31b_v1 · rescored 77adc97b: 2.154triage_hosted_gemma4_31b_v1 · rescored cd35dc1f: 2.154triage_local_x3_qwen38_27b_v2: 2.2triage_local_x3_qwen38_27b_v2 · rescored 3d3a6afd: 2.2triage_local_x3_qwen38_27b_v2 · rescored 5f2c057c: 2.2triage_deepseek_41flash_v1 · rescored 5f2c057c: 2.246triage_deepseek_41flash_v1 · rescored 77adc97b: 2.246triage_deepseek_41flash_v1 · rescored cd35dc1f: 2.246triage_deepseek_41flash_v2: 2.262triage_local_dgx_ministral_3_14b_v1: 2.262triage_deepseek_41flash_v2 · rescored 3d3a6afd: 2.262triage_local_dgx_ministral_3_14b_v1 · rescored 3d3a6afd: 2.262triage_deepseek_41flash_v2 · rescored 5f2c057c: 2.262triage_local_dgx_ministral_3_14b_v1 · rescored 5f2c057c: 2.262triage_local_dgx_ministral_3_14b_v1 · rescored 77adc97b: 2.262triage_local_dgx_ministral_3_14b_v1 · rescored cd35dc1f: 2.262triage_local_dgx_qwen38_27b_v2: 2.277triage_local_mac_qwen38_27b_v1: 2.277triage_local_dgx_qwen38_27b_v2 · rescored 3d3a6afd: 2.277triage_local_dgx_qwen38_27b_v2 · rescored 5f2c057c: 2.277triage_local_mac_qwen38_27b_v1 · rescored 5f2c057c: 2.277triage_local_mac_qwen38_27b_v1 · rescored 77adc97b: 2.277triage_local_mac_qwen38_27b_v1 · rescored cd35dc1f: 2.277triage_baseline_pro_noreason_v1: 2.292triage_local_x3_qwen38_27b_v1: 2.292triage_baseline_pro_noreason_v1 · rescored 3d3a6afd: 2.292triage_baseline_pro_noreason_v1 · rescored 5f2c057c: 2.292triage_local_x3_qwen38_27b_v1 · rescored 5f2c057c: 2.292triage_local_x3_qwen38_27b_v1 · rescored 77adc97b: 2.292triage_local_x3_qwen38_27b_v1 · rescored cd35dc1f: 2.292triage_local_dgx_gemma4_31b_v2: 2.308triage_local_mac_qwen38_27b_v2: 2.308triage_local_dgx_gemma4_31b_v2 · rescored 3d3a6afd: 2.308triage_local_mac_qwen38_27b_v2 · rescored 3d3a6afd: 2.308triage_local_dgx_gemma4_31b_v2 · rescored 5f2c057c: 2.308triage_local_mac_qwen38_27b_v2 · rescored 5f2c057c: 2.308triage_local_dgx_qwen38_27b_v1: 2.354triage_local_dgx_qwen38_27b_v1 · rescored 5f2c057c: 2.354triage_local_dgx_qwen38_27b_v1 · rescored 77adc97b: 2.354triage_local_dgx_qwen38_27b_v1 · rescored cd35dc1f: 2.354triage_local_dgx_qwen35_122b_v2: 2.369triage_local_x3_gemma4_31b_v2: 2.369triage_local_dgx_qwen35_122b_v2 · rescored 3d3a6afd: 2.369triage_local_x3_gemma4_31b_v2 · rescored 3d3a6afd: 2.369triage_local_dgx_qwen35_122b_v2 · rescored 5f2c057c: 2.369triage_local_x3_gemma4_31b_v2 · rescored 5f2c057c: 2.369triage_hosted_gemma4_31b_v2: 2.385triage_hosted_gemma4_31b_v2 · rescored 3d3a6afd: 2.385triage_hosted_gemma4_31b_v2 · rescored 5f2c057c: 2.385triage_local_mac_gemma4_31b_v2: 2.4triage_local_mac_gemma4_31b_v2 · rescored 3d3a6afd: 2.4triage_local_mac_gemma4_31b_v2 · rescored 5f2c057c: 2.4triage_local_mac_ministral_3_14b_v1: 2.415triage_local_mac_ministral_3_14b_v1 · rescored cd35dc1f: 2.415triage_local_mac_ministral_3_14b_v1 · rescored 3d3a6afd: 2.492triage_local_mac_ministral_3_14b_v1 · rescored 5f2c057c: 2.492triage_local_mac_ministral_3_14b_v1 · rescored 77adc97b: 2.492triage_local_x3_ministral_3_14b_v1: 2.508triage_local_x3_ministral_3_14b_v1 · rescored 3d3a6afd: 2.508triage_local_x3_ministral_3_14b_v1 · rescored 5f2c057c: 2.508triage_local_x3_ministral_3_14b_v1 · rescored 77adc97b: 2.508triage_local_x3_ministral_3_14b_v1 · rescored cd35dc1f: 2.508triage_stub_dismiss_v1: 2.615triage_stub_dismiss_v1 · rescored 3d3a6afd: 2.615triage_stub_dismiss_v1 · rescored 5f2c057c: 2.615triage_stub_dismiss_v1 · rescored 77adc97b: 2.615triage_stub_dismiss_v1 · rescored cd35dc1f: 2.6150.477 · better2.615

scored against silverscored by a judgewrote the reference it is scored againstscored against a label set, not a reference setunmarked rows are scored against gold

armmean model_triage_costitemsscored against
triage_stub_escalate_v10.47765gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored 3d3a6afd0.47765gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored 5f2c057c0.47765gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored 77adc97b0.47765gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored cd35dc1f0.47765gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 3d3a6afd, repeats ×30.52365gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 5f2c057c, repeats ×30.52365gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 77adc97b, repeats ×30.52365gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored cd35dc1f, repeats ×30.52365gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 3d3a6afd0.55465gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 5f2c057c0.55465gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 77adc97b0.55465gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored cd35dc1f0.55465gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 5f2c057c0.61565gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 77adc97b0.61565gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored cd35dc1f0.61565gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 5f2c057c, repeats ×30.68265gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 77adc97b, repeats ×30.68265gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored cd35dc1f, repeats ×30.68265gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v10.78565gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v10.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored 3d3a6afd0.78565gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored 3d3a6afd0.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored 5f2c057c0.78565gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored 5f2c057c0.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored 77adc97b0.78565gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored 77adc97b0.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored cd35dc1f0.78565gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored cd35dc1f0.78565gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v20.83165gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v2 · rescored 3d3a6afd0.83165gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v2 · rescored 5f2c057c0.83165gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v10.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored 3d3a6afd0.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored 5f2c057c0.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored 77adc97b0.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored cd35dc1f0.86265gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 3d3a6afd0.93865gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 5f2c057c0.93865gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 77adc97b0.93865gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored cd35dc1f0.93865gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v10.96965gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 3d3a6afd, repeats ×30.96965gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v1 · rescored 5f2c057c0.96965gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 5f2c057c, repeats ×30.96965gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v1 · rescored 77adc97b0.96965gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 77adc97b, repeats ×30.96965gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v1 · rescored cd35dc1f0.96965gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v1 · rescored cd35dc1f0.96965gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored cd35dc1f, repeats ×30.96965gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v11.01565gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · repeats1.01565gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c1.01565gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c, repeats1.01565gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b1.01565gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b, repeats1.01565gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f1.01565gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f, repeats1.01565gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v11.04665gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 5f2c057c1.04665gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 77adc97b1.04665gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v1 · rescored cd35dc1f1.04665gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v11.06265gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored 3d3a6afd1.06265gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored 3d3a6afd1.06265gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored 5f2c057c1.06265gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored 5f2c057c1.06265gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored 77adc97b1.06265gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored 77adc97b1.06265gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored cd35dc1f1.06265gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored cd35dc1f1.06265gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v11.09265gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored 3d3a6afd1.09265gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored 5f2c057c1.09265gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored 77adc97b1.09265gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored cd35dc1f1.09265gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v11.13865gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v11.13865gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 3d3a6afd1.13865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 5f2c057c1.13865gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 5f2c057c1.13865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 77adc97b1.13865gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 77adc97b1.13865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored cd35dc1f1.13865gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored cd35dc1f1.13865gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored 3d3a6afd1.18565gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored 5f2c057c1.18565gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored 77adc97b1.18565gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored cd35dc1f1.18565gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v11.24665gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c1.24665gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 77adc97b1.24665gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f1.24665gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v11.27765gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v1 · rescored cd35dc1f1.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v11.29265gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c1.29265gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 77adc97b1.29265gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f1.29265gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v11.43165gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v1 · rescored cd35dc1f1.43165gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v11.44665gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 5f2c057c1.44665gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 77adc97b1.44665gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v1 · rescored cd35dc1f1.44665gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v21.46265gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v21.46265gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v21.46265gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v2 · rescored 3d3a6afd1.46265gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd1.46265gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd1.46265gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v2 · rescored 5f2c057c1.46265gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 5f2c057c1.46265gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 5f2c057c1.46265gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v21.47765gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v2 · rescored 3d3a6afd1.47765gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v2 · rescored 5f2c057c1.47765gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored 3d3a6afd1.49265gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored 5f2c057c1.49265gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored 77adc97b1.49265gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored cd35dc1f1.49265gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v11.50865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v11.50865gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored 3d3a6afd1.50865gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored 5f2c057c1.50865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v1 · rescored 5f2c057c1.50865gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored 77adc97b1.50865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v1 · rescored 77adc97b1.50865gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored cd35dc1f1.50865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v1 · rescored cd35dc1f1.50865gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v11.52365gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v21.52365gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v2 · rescored 3d3a6afd1.52365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v1 · rescored 5f2c057c1.52365gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v2 · rescored 5f2c057c1.52365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v1 · rescored 77adc97b1.52365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v1 · rescored cd35dc1f1.52365gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v21.53865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd1.53865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 5f2c057c1.53865gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v21.55465gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v21.55465gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 3d3a6afd1.55465gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v2 · rescored 3d3a6afd1.55465gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd1.55465gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 5f2c057c1.55465gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v2 · rescored 5f2c057c1.55465gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c1.55465gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 77adc97b1.55465gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored cd35dc1f1.55465gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v11.56965gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored 3d3a6afd1.56965gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored 5f2c057c1.56965gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored 77adc97b1.56965gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored cd35dc1f1.56965gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v11.60065gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored 3d3a6afd1.60065gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored 5f2c057c1.60065gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored 77adc97b1.60065gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored cd35dc1f1.60065gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v11.63165gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v1 · rescored 5f2c057c1.63165gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v1 · rescored 77adc97b1.63165gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v1 · rescored cd35dc1f1.63165gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v11.64665gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v1 · rescored 5f2c057c1.64665gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v1 · rescored 5f2c057c1.64665gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v1 · rescored 77adc97b1.64665gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v1 · rescored 77adc97b1.64665gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v1 · rescored cd35dc1f1.64665gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v1 · rescored cd35dc1f1.64665gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v11.66265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v11.66265gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored 3d3a6afd1.66265gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored 3d3a6afd1.66265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored 3d3a6afd1.66265gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored 5f2c057c1.66265gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored 5f2c057c1.66265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored 5f2c057c1.66265gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v1 · rescored 5f2c057c1.66265gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored 77adc97b1.66265gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored 77adc97b1.66265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored 77adc97b1.66265gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v1 · rescored 77adc97b1.66265gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored cd35dc1f1.66265gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored cd35dc1f1.66265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored cd35dc1f1.66265gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v1 · rescored 5f2c057c1.69265gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v1 · rescored 77adc97b1.69265gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v1 · rescored cd35dc1f1.69265gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v21.70865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd1.70865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c1.70865gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v21.72365gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v11.72365gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v21.72365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v2 · rescored 3d3a6afd1.72365gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd1.72365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v2 · rescored 5f2c057c1.72365gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v1 · rescored 5f2c057c1.72365gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c1.72365gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v1 · rescored 77adc97b1.72365gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v1 · rescored cd35dc1f1.72365gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v21.73865gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v2 · rescored 3d3a6afd1.73865gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v2 · rescored 5f2c057c1.73865gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 3d3a6afd, repeats ×31.76465gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 5f2c057c, repeats ×31.76465gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 77adc97b, repeats ×31.76465gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored cd35dc1f, repeats ×31.76465gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v21.78565gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v2 · rescored 3d3a6afd1.78565gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v2 · rescored 5f2c057c1.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v11.80065gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v1 · rescored 5f2c057c1.80065gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v1 · rescored 77adc97b1.80065gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v1 · rescored cd35dc1f1.80065gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v1 · rescored 5f2c057c1.81565gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v1 · rescored 77adc97b1.81565gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v11.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v11.86265gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored 3d3a6afd1.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored 3d3a6afd1.86265gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored 5f2c057c1.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored 5f2c057c1.86265gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored 77adc97b1.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored 77adc97b1.86265gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored cd35dc1f1.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored cd35dc1f1.86265gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v21.90865gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v2 · rescored 3d3a6afd1.90865gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v2 · rescored 5f2c057c1.90865gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v1 · rescored 5f2c057c1.92365gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v1 · rescored 77adc97b1.92365gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v1 · rescored cd35dc1f1.92365gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v1 · rescored 5f2c057c1.96965gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v1 · rescored 77adc97b1.96965gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v21.98565gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v2 · rescored 3d3a6afd1.98565gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v2 · rescored 5f2c057c1.98565gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored 3d3a6afd2.03165gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored 5f2c057c2.03165gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored 77adc97b2.03165gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored cd35dc1f2.03165gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v12.10865gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored 3d3a6afd2.10865gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored 5f2c057c2.10865gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored 77adc97b2.10865gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored cd35dc1f2.10865gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v1 · rescored 5f2c057c2.15465gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v1 · rescored 77adc97b2.15465gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v1 · rescored cd35dc1f2.15465gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v22.20065gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v2 · rescored 3d3a6afd2.20065gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v2 · rescored 5f2c057c2.20065gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v1 · rescored 5f2c057c2.24665gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v1 · rescored 77adc97b2.24665gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v1 · rescored cd35dc1f2.24665gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v22.26265gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v12.26265gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v2 · rescored 3d3a6afd2.26265gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored 3d3a6afd2.26265gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v2 · rescored 5f2c057c2.26265gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored 5f2c057c2.26265gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored 77adc97b2.26265gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored cd35dc1f2.26265gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v22.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v12.27765gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v2 · rescored 3d3a6afd2.27765gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v2 · rescored 5f2c057c2.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v1 · rescored 5f2c057c2.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v1 · rescored 77adc97b2.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v1 · rescored cd35dc1f2.27765gold/orrery_triage_v1_labelled gold
triage_baseline_pro_noreason_v12.29265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v12.29265gold/orrery_triage_v1_labelled gold
triage_baseline_pro_noreason_v1 · rescored 3d3a6afd2.29265gold/orrery_triage_v1_labelled gold
triage_baseline_pro_noreason_v1 · rescored 5f2c057c2.29265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v1 · rescored 5f2c057c2.29265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v1 · rescored 77adc97b2.29265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v1 · rescored cd35dc1f2.29265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v22.30865gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v22.30865gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v2 · rescored 3d3a6afd2.30865gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v2 · rescored 3d3a6afd2.30865gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v2 · rescored 5f2c057c2.30865gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v2 · rescored 5f2c057c2.30865gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v12.35465gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v1 · rescored 5f2c057c2.35465gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v1 · rescored 77adc97b2.35465gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v1 · rescored cd35dc1f2.35465gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v22.36965gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v22.36965gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v2 · rescored 3d3a6afd2.36965gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v2 · rescored 3d3a6afd2.36965gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v2 · rescored 5f2c057c2.36965gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v2 · rescored 5f2c057c2.36965gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v22.38565gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v2 · rescored 3d3a6afd2.38565gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v2 · rescored 5f2c057c2.38565gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v22.40065gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v2 · rescored 3d3a6afd2.40065gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v2 · rescored 5f2c057c2.40065gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v12.41565gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored cd35dc1f2.41565gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored 3d3a6afd2.49265gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored 5f2c057c2.49265gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored 77adc97b2.49265gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v12.50865gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored 3d3a6afd2.50865gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored 5f2c057c2.50865gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored 77adc97b2.50865gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored cd35dc1f2.50865gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v12.61565gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored 3d3a6afd2.61565gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored 5f2c057c2.61565gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored 77adc97b2.61565gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored cd35dc1f2.61565gold/orrery_triage_v1_labelled gold