eval-harness

metric

triage_cost_p10

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 with a missed real problem priced at 10 false alarms instead of 5.

A project explains its own metric at length with METRIC_DESCRIPTIONS = {"triage_cost_p10": {"summary": "…", "details": "…"}}.

02on orrery_triage_v1_labelled

gold/orrery_triage_v1_labelledtriage_local_dgx_llama31_8b_v1: 0.415triage_local_mac_llama31_8b_v1: 0.415triage_local_dgx_llama31_8b_v1 · rescored 3d3a6afd: 0.415triage_local_mac_llama31_8b_v1 · rescored 3d3a6afd: 0.415triage_local_x3_llama31_8b_v1 · rescored 3d3a6afd: 0.415triage_local_dgx_llama31_8b_v1 · rescored 5f2c057c: 0.415triage_local_mac_llama31_8b_v1 · rescored 5f2c057c: 0.415triage_local_x3_llama31_8b_v1 · rescored 5f2c057c: 0.415triage_local_dgx_llama31_8b_v1 · rescored 77adc97b: 0.415triage_local_mac_llama31_8b_v1 · rescored 77adc97b: 0.415triage_local_x3_llama31_8b_v1 · rescored 77adc97b: 0.415triage_local_dgx_llama31_8b_v1 · rescored cd35dc1f: 0.415triage_local_mac_llama31_8b_v1 · rescored cd35dc1f: 0.415triage_local_x3_llama31_8b_v1 · rescored cd35dc1f: 0.415triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1: 0.477triage_local_x3_gpt_oss_20b_v1: 0.477triage_stub_escalate_v1: 0.477triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 3d3a6afd: 0.477triage_stub_escalate_v1 · rescored 3d3a6afd: 0.477triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 5f2c057c: 0.477triage_stub_escalate_v1 · rescored 5f2c057c: 0.477triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 77adc97b: 0.477triage_stub_escalate_v1 · rescored 77adc97b: 0.477triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored cd35dc1f: 0.477triage_local_x3_gpt_oss_20b_v1 · rescored cd35dc1f: 0.477triage_stub_escalate_v1 · rescored cd35dc1f: 0.477triage_local_dgx_gpt_oss_20b_v1: 0.631triage_local_dgx_gpt_oss_20b_v1 · rescored cd35dc1f: 0.631triage_local_dgx_nemotron_35_bolt_e4b_v1: 0.754triage_local_x3_nemotron_35_lightning_30b_v1: 0.754triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c: 0.754triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 5f2c057c: 0.754triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 77adc97b: 0.754triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 77adc97b: 0.754triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f: 0.754triage_local_x3_nemotron_35_lightning_30b_v1 · rescored cd35dc1f: 0.754triage_gemini_35_flash_lite_v1 · rescored 3d3a6afd: 0.769triage_gemini_35_flash_lite_v1 · rescored 5f2c057c: 0.769triage_gemini_35_flash_lite_v1 · rescored 77adc97b: 0.769triage_gemini_35_flash_lite_v1 · rescored cd35dc1f: 0.769triage_gpt_6_luna_v2: 0.785triage_local_mac_gpt_oss_20b_v1: 0.785triage_gpt_6_luna_v2 · rescored 3d3a6afd: 0.785triage_gpt_6_luna_v2 · rescored 5f2c057c: 0.785triage_local_mac_gpt_oss_20b_v1 · rescored cd35dc1f: 0.785triage_gpt_61_sol_v1 · rescored 3d3a6afd, repeats ×3: 0.815triage_gpt_6_luna_v1 · rescored 5f2c057c: 0.815triage_gpt_61_sol_v1 · rescored 5f2c057c, repeats ×3: 0.815triage_gpt_6_luna_v1 · rescored 77adc97b: 0.815triage_gpt_61_sol_v1 · rescored 77adc97b, repeats ×3: 0.815triage_gpt_6_luna_v1 · rescored cd35dc1f: 0.815triage_gpt_61_sol_v1 · rescored cd35dc1f, repeats ×3: 0.815triage_local_x3_nemotron_35_bolt_e4b_v2: 0.862triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd: 0.862triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c: 0.862triage_local_dgx_gemma4_12b_noreason_v1: 0.892triage_local_mac_gemma4_12b_noreason_v1: 0.892triage_local_mac_nemotron_35_bolt_e4b_v2: 0.892triage_haiku_55_v1 · rescored 3d3a6afd: 0.892triage_local_dgx_gemma4_12b_noreason_v1 · rescored 3d3a6afd: 0.892triage_local_mac_gemma4_12b_noreason_v1 · rescored 3d3a6afd: 0.892triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd: 0.892triage_haiku_55_v1 · rescored 5f2c057c: 0.892triage_local_dgx_gemma4_12b_noreason_v1 · rescored 5f2c057c: 0.892triage_local_mac_gemma4_12b_noreason_v1 · rescored 5f2c057c: 0.892triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c: 0.892triage_haiku_55_v1 · rescored 77adc97b: 0.892triage_local_dgx_gemma4_12b_noreason_v1 · rescored 77adc97b: 0.892triage_local_mac_gemma4_12b_noreason_v1 · rescored 77adc97b: 0.892triage_haiku_55_v1 · rescored cd35dc1f: 0.892triage_local_dgx_gemma4_12b_noreason_v1 · rescored cd35dc1f: 0.892triage_local_mac_gemma4_12b_noreason_v1 · rescored cd35dc1f: 0.892triage_local_dgx_nemotron_35_lightning_30b_v2: 0.908triage_local_x3_nemotron_35_bolt_e4b_v1: 0.908triage_local_x3_nemotron_35_lightning_30b_v2: 0.908triage_local_x3_nemotron_35_bolt_e4b_v1 · repeats: 0.908triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd: 0.908triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd: 0.908triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 5f2c057c: 0.908triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c: 0.908triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 5f2c057c: 0.908triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c, repeats: 0.908triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b: 0.908triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b, repeats: 0.908triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f: 0.908triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f, repeats: 0.908triage_haiku_55_v1 · rescored 3d3a6afd, repeats ×3: 0.918triage_haiku_55_v1 · rescored 5f2c057c, repeats ×3: 0.918triage_haiku_55_v1 · rescored 77adc97b, repeats ×3: 0.918triage_haiku_55_v1 · rescored cd35dc1f, repeats ×3: 0.918triage_local_mac_nemotron_35_lightning_30b_v2: 0.923triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd: 0.923triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 5f2c057c: 0.923triage_gpt_6_luna_v1 · rescored 5f2c057c, repeats ×3: 0.933triage_gpt_6_luna_v1 · rescored 77adc97b, repeats ×3: 0.933triage_gpt_6_luna_v1 · rescored cd35dc1f, repeats ×3: 0.933triage_hosted_gpt_oss_120b_v2: 0.938triage_hosted_gpt_oss_120b_v2 · rescored 3d3a6afd: 0.938triage_hosted_gpt_oss_120b_v2 · rescored 5f2c057c: 0.938triage_local_dgx_nemotron_35_bolt_e4b_v2: 0.954triage_local_dgx_nemotron_35_lightning_30b_v1: 0.954triage_local_mac_nemotron_35_lightning_30b_v1: 0.954triage_gemini_4_argon_v1 · rescored 3d3a6afd: 0.954triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd: 0.954triage_gemini_4_argon_v1 · rescored 5f2c057c: 0.954triage_kimi_k31_pro_v1 · rescored 5f2c057c: 0.954triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c: 0.954triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 5f2c057c: 0.954triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 5f2c057c: 0.954triage_gemini_4_argon_v1 · rescored 77adc97b: 0.954triage_kimi_k31_pro_v1 · rescored 77adc97b: 0.954triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 77adc97b: 0.954triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 77adc97b: 0.954triage_gemini_4_argon_v1 · rescored cd35dc1f: 0.954triage_kimi_k31_pro_v1 · rescored cd35dc1f: 0.954triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored cd35dc1f: 0.954triage_local_mac_nemotron_35_lightning_30b_v1 · rescored cd35dc1f: 0.954triage_local_dgx_phi4_14b_v1: 0.969triage_gpt_61_sol_v1 · rescored 3d3a6afd: 0.969triage_local_dgx_phi4_14b_v1 · rescored 3d3a6afd: 0.969triage_gpt_61_sol_v1 · rescored 5f2c057c: 0.969triage_local_dgx_phi4_14b_v1 · rescored 5f2c057c: 0.969triage_gpt_61_sol_v1 · rescored 77adc97b: 0.969triage_local_dgx_phi4_14b_v1 · rescored 77adc97b: 0.969triage_gpt_61_sol_v1 · rescored cd35dc1f: 0.969triage_local_dgx_phi4_14b_v1 · rescored cd35dc1f: 0.969triage_local_dgx_nemotron_35_super_120b_v2: 0.985triage_local_dgx_nemotron_35_super_120b_v2 · rescored 3d3a6afd: 0.985triage_local_dgx_nemotron_35_super_120b_v2 · rescored 5f2c057c: 0.985triage_local_dgx_gpt_oss_20b_v2: 1.031triage_local_dgx_gpt_oss_20b_v2 · rescored 3d3a6afd: 1.031triage_local_dgx_gpt_oss_20b_v2 · rescored 5f2c057c: 1.031triage_local_dgx_gpt_oss_120b_v2: 1.046triage_local_dgx_gpt_oss_120b_v2 · rescored 3d3a6afd: 1.046triage_local_dgx_gpt_oss_120b_v2 · rescored 5f2c057c: 1.046triage_local_dgx_gemma4_31b_v1: 1.062triage_local_dgx_nemotron_35_super_120b_v1: 1.062triage_local_mac_gemma4_31b_v1: 1.062triage_hosted_gpt_oss_120b_v1 · rescored 5f2c057c: 1.062triage_local_dgx_gemma4_31b_v1 · rescored 5f2c057c: 1.062triage_local_dgx_nemotron_35_super_120b_v1 · rescored 5f2c057c: 1.062triage_local_mac_gemma4_31b_v1 · rescored 5f2c057c: 1.062triage_hosted_gpt_oss_120b_v1 · rescored 77adc97b: 1.062triage_local_dgx_gemma4_31b_v1 · rescored 77adc97b: 1.062triage_local_dgx_nemotron_35_super_120b_v1 · rescored 77adc97b: 1.062triage_local_mac_gemma4_31b_v1 · rescored 77adc97b: 1.062triage_hosted_gpt_oss_120b_v1 · rescored cd35dc1f: 1.062triage_local_dgx_gemma4_31b_v1 · rescored cd35dc1f: 1.062triage_local_dgx_nemotron_35_super_120b_v1 · rescored cd35dc1f: 1.062triage_local_mac_gemma4_31b_v1 · rescored cd35dc1f: 1.062triage_local_dgx_gpt_oss_120b_v1: 1.077triage_local_x3_gemma4_31b_v1: 1.077triage_local_dgx_gpt_oss_120b_v1 · rescored 5f2c057c: 1.077triage_local_x3_gemma4_31b_v1 · rescored 5f2c057c: 1.077triage_local_dgx_gpt_oss_120b_v1 · rescored 77adc97b: 1.077triage_local_x3_gemma4_31b_v1 · rescored 77adc97b: 1.077triage_local_dgx_gpt_oss_120b_v1 · rescored cd35dc1f: 1.077triage_local_x3_gemma4_31b_v1 · rescored cd35dc1f: 1.077triage_local_mac_gpt_oss_20b_v2: 1.092triage_local_mac_gpt_oss_20b_v2 · rescored 3d3a6afd: 1.092triage_sonnet_55_v1 · rescored 3d3a6afd: 1.092triage_local_mac_gpt_oss_20b_v2 · rescored 5f2c057c: 1.092triage_sonnet_55_v1 · rescored 5f2c057c: 1.092triage_sonnet_55_v1 · rescored 77adc97b: 1.092triage_sonnet_55_v1 · rescored cd35dc1f: 1.092triage_kimi_k31_pro_v2: 1.108triage_kimi_k31_pro_v2 · rescored 3d3a6afd: 1.108triage_kimi_k31_pro_v2 · rescored 5f2c057c: 1.108triage_local_mac_phi4_14b_v1: 1.138triage_local_mac_phi4_14b_v1 · rescored 3d3a6afd: 1.138triage_local_mac_phi4_14b_v1 · rescored 5f2c057c: 1.138triage_local_mac_phi4_14b_v1 · rescored 77adc97b: 1.138triage_local_mac_phi4_14b_v1 · rescored cd35dc1f: 1.138triage_local_x3_phi4_14b_v1: 1.154triage_local_x3_phi4_14b_v1 · rescored 3d3a6afd: 1.154triage_local_x3_phi4_14b_v1 · rescored 5f2c057c: 1.154triage_local_x3_phi4_14b_v1 · rescored 77adc97b: 1.154triage_local_x3_phi4_14b_v1 · rescored cd35dc1f: 1.154triage_hosted_qwen3_30b_a3b_v1: 1.169triage_local_x3_gemma4_31b_v2: 1.169triage_hosted_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 1.169triage_local_x3_gemma4_31b_v2 · rescored 3d3a6afd: 1.169triage_hosted_qwen3_30b_a3b_v1 · rescored 5f2c057c: 1.169triage_local_x3_gemma4_31b_v2 · rescored 5f2c057c: 1.169triage_hosted_qwen3_30b_a3b_v1 · rescored 77adc97b: 1.169triage_hosted_qwen3_30b_a3b_v1 · rescored cd35dc1f: 1.169triage_hosted_gemma4_31b_v2: 1.185triage_local_dgx_gemma4_31b_v2: 1.185triage_hosted_gemma4_31b_v2 · rescored 3d3a6afd: 1.185triage_local_dgx_gemma4_31b_v2 · rescored 3d3a6afd: 1.185triage_hosted_gemma4_31b_v1 · rescored 5f2c057c: 1.185triage_hosted_gemma4_31b_v2 · rescored 5f2c057c: 1.185triage_local_dgx_gemma4_31b_v2 · rescored 5f2c057c: 1.185triage_hosted_gemma4_31b_v1 · rescored 77adc97b: 1.185triage_hosted_gemma4_31b_v1 · rescored cd35dc1f: 1.185triage_baseline_pro_v1 · rescored 3d3a6afd, repeats ×3: 1.2triage_baseline_pro_v1 · rescored 5f2c057c, repeats ×3: 1.2triage_baseline_pro_v1 · rescored 77adc97b, repeats ×3: 1.2triage_baseline_pro_v1 · rescored cd35dc1f, repeats ×3: 1.2triage_local_mac_nemotron_35_bolt_e4b_v1: 1.215triage_deepseek_41flash_v1 · rescored 5f2c057c: 1.215triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c: 1.215triage_deepseek_41flash_v1 · rescored 77adc97b: 1.215triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 77adc97b: 1.215triage_deepseek_41flash_v1 · rescored cd35dc1f: 1.215triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f: 1.215triage_baseline_pro_v1 · rescored 3d3a6afd: 1.246triage_baseline_pro_v1 · rescored 5f2c057c: 1.246triage_baseline_pro_v1 · rescored 77adc97b: 1.246triage_baseline_pro_v1 · rescored cd35dc1f: 1.246triage_local_x3_qwen3_30b_a3b_v1: 1.262triage_mistral_large_4_v2: 1.262triage_local_x3_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 1.262triage_mistral_large_4_v2 · rescored 3d3a6afd: 1.262triage_local_x3_qwen3_30b_a3b_v1 · rescored 5f2c057c: 1.262triage_mistral_large_4_v1 · rescored 5f2c057c: 1.262triage_mistral_large_4_v2 · rescored 5f2c057c: 1.262triage_local_x3_qwen3_30b_a3b_v1 · rescored 77adc97b: 1.262triage_mistral_large_4_v1 · rescored 77adc97b: 1.262triage_local_x3_qwen3_30b_a3b_v1 · rescored cd35dc1f: 1.262triage_mistral_large_4_v1 · rescored cd35dc1f: 1.262triage_local_x3_gpt_oss_20b_v2: 1.277triage_local_x3_gpt_oss_20b_v2 · rescored 3d3a6afd: 1.277triage_local_x3_gpt_oss_20b_v2 · rescored 5f2c057c: 1.277triage_local_mac_gemma4_31b_v2: 1.338triage_local_mac_gemma4_31b_v2 · rescored 3d3a6afd: 1.338triage_local_mac_gemma4_31b_v2 · rescored 5f2c057c: 1.338triage_local_x3_ministral_3_14b_v1: 1.354triage_local_x3_ministral_3_14b_v1 · rescored 3d3a6afd: 1.354triage_local_x3_ministral_3_14b_v1 · rescored 5f2c057c: 1.354triage_local_x3_ministral_3_14b_v1 · rescored 77adc97b: 1.354triage_local_x3_ministral_3_14b_v1 · rescored cd35dc1f: 1.354triage_local_dgx_qwen38_27b_v1: 1.369triage_local_mac_ministral_3_14b_v1: 1.369triage_local_dgx_qwen38_27b_v1 · rescored 5f2c057c: 1.369triage_local_dgx_qwen38_27b_v1 · rescored 77adc97b: 1.369triage_local_dgx_qwen38_27b_v1 · rescored cd35dc1f: 1.369triage_local_mac_ministral_3_14b_v1 · rescored cd35dc1f: 1.369triage_stub_dismiss_v1: 1.4triage_stub_dismiss_v1 · rescored 3d3a6afd: 1.4triage_stub_dismiss_v1 · rescored 5f2c057c: 1.4triage_stub_dismiss_v1 · rescored 77adc97b: 1.4triage_stub_dismiss_v1 · rescored cd35dc1f: 1.4triage_local_dgx_ministral_3_14b_v1: 1.446triage_local_mac_qwen3_30b_a3b_v1: 1.446triage_local_dgx_ministral_3_14b_v1 · rescored 3d3a6afd: 1.446triage_local_mac_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 1.446triage_local_dgx_ministral_3_14b_v1 · rescored 5f2c057c: 1.446triage_local_mac_qwen3_30b_a3b_v1 · rescored 5f2c057c: 1.446triage_local_dgx_ministral_3_14b_v1 · rescored 77adc97b: 1.446triage_local_mac_qwen3_30b_a3b_v1 · rescored 77adc97b: 1.446triage_local_dgx_ministral_3_14b_v1 · rescored cd35dc1f: 1.446triage_local_mac_qwen3_30b_a3b_v1 · rescored cd35dc1f: 1.446triage_hosted_qwen35_122b_v2: 1.492triage_hosted_mistral_medium_35_v1 · rescored 3d3a6afd: 1.492triage_hosted_qwen35_122b_v2 · rescored 3d3a6afd: 1.492triage_hosted_mistral_medium_35_v1 · rescored 5f2c057c: 1.492triage_hosted_qwen35_122b_v2 · rescored 5f2c057c: 1.492triage_hosted_mistral_medium_35_v1 · rescored 77adc97b: 1.492triage_hosted_mistral_medium_35_v1 · rescored cd35dc1f: 1.492triage_deepseek_41flash_v2: 1.523triage_local_x3_qwen38_27b_v1: 1.523triage_deepseek_41flash_v2 · rescored 3d3a6afd: 1.523triage_local_mac_ministral_3_14b_v1 · rescored 3d3a6afd: 1.523triage_deepseek_41flash_v2 · rescored 5f2c057c: 1.523triage_local_mac_ministral_3_14b_v1 · rescored 5f2c057c: 1.523triage_local_x3_qwen38_27b_v1 · rescored 5f2c057c: 1.523triage_local_mac_ministral_3_14b_v1 · rescored 77adc97b: 1.523triage_local_x3_qwen38_27b_v1 · rescored 77adc97b: 1.523triage_local_x3_qwen38_27b_v1 · rescored cd35dc1f: 1.523triage_local_dgx_qwen3_30b_a3b_v1: 1.585triage_local_dgx_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 1.585triage_local_dgx_qwen3_30b_a3b_v1 · rescored 5f2c057c: 1.585triage_local_dgx_qwen3_30b_a3b_v1 · rescored 77adc97b: 1.585triage_local_dgx_qwen3_30b_a3b_v1 · rescored cd35dc1f: 1.585triage_local_dgx_mistral_medium_35_128b_v1: 1.692triage_local_dgx_mistral_medium_35_128b_v1 · rescored 3d3a6afd: 1.692triage_local_dgx_mistral_medium_35_128b_v1 · rescored 5f2c057c: 1.692triage_local_dgx_mistral_medium_35_128b_v1 · rescored 77adc97b: 1.692triage_local_dgx_mistral_medium_35_128b_v1 · rescored cd35dc1f: 1.692triage_local_dgx_gpt_oss_20b_v1 · rescored 5f2c057c: 1.708triage_local_dgx_gpt_oss_20b_v1 · rescored 77adc97b: 1.708triage_local_mac_qwen38_27b_v1: 1.815triage_local_mac_qwen38_27b_v1 · rescored 5f2c057c: 1.815triage_local_mac_qwen38_27b_v1 · rescored 77adc97b: 1.815triage_local_mac_qwen38_27b_v1 · rescored cd35dc1f: 1.815triage_hosted_qwen35_122b_v1 · rescored 5f2c057c: 1.846triage_hosted_qwen35_122b_v1 · rescored 77adc97b: 1.846triage_hosted_qwen35_122b_v1 · rescored cd35dc1f: 1.846triage_local_mac_gpt_oss_20b_v1 · rescored 5f2c057c: 1.862triage_local_x3_gpt_oss_20b_v1 · rescored 5f2c057c: 1.862triage_local_mac_gpt_oss_20b_v1 · rescored 77adc97b: 1.862triage_local_x3_gpt_oss_20b_v1 · rescored 77adc97b: 1.862triage_local_dgx_qwen35_122b_v1: 1.985triage_local_dgx_qwen38_27b_v2: 1.985triage_local_mac_qwen38_27b_v2: 1.985triage_local_dgx_qwen38_27b_v2 · rescored 3d3a6afd: 1.985triage_local_mac_qwen38_27b_v2 · rescored 3d3a6afd: 1.985triage_local_dgx_qwen35_122b_v1 · rescored 5f2c057c: 1.985triage_local_dgx_qwen38_27b_v2 · rescored 5f2c057c: 1.985triage_local_mac_qwen38_27b_v2 · rescored 5f2c057c: 1.985triage_local_dgx_qwen35_122b_v1 · rescored 77adc97b: 1.985triage_local_dgx_qwen35_122b_v1 · rescored cd35dc1f: 1.985triage_baseline_pro_noreason_v1: 2.077triage_baseline_pro_noreason_v1 · rescored 3d3a6afd: 2.077triage_baseline_pro_noreason_v1 · rescored 5f2c057c: 2.077triage_local_dgx_llama4_scout_v1: 2.108triage_local_dgx_llama4_scout_v1 · rescored 3d3a6afd: 2.108triage_local_dgx_llama4_scout_v1 · rescored 5f2c057c: 2.108triage_local_dgx_llama4_scout_v1 · rescored 77adc97b: 2.108triage_local_dgx_llama4_scout_v1 · rescored cd35dc1f: 2.108triage_local_x3_qwen38_27b_v2: 2.292triage_local_x3_qwen38_27b_v2 · rescored 3d3a6afd: 2.292triage_local_x3_qwen38_27b_v2 · rescored 5f2c057c: 2.292triage_local_dgx_qwen35_122b_v2: 2.431triage_local_dgx_qwen35_122b_v2 · rescored 3d3a6afd: 2.431triage_local_dgx_qwen35_122b_v2 · rescored 5f2c057c: 2.4310.415 · better2.431

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 triage_cost_p10itemsscored against
triage_local_dgx_llama31_8b_v10.41565gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v10.41565gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored 3d3a6afd0.41565gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored 3d3a6afd0.41565gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored 3d3a6afd0.41565gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored 5f2c057c0.41565gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored 5f2c057c0.41565gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored 5f2c057c0.41565gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored 77adc97b0.41565gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored 77adc97b0.41565gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored 77adc97b0.41565gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored cd35dc1f0.41565gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored cd35dc1f0.41565gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored cd35dc1f0.41565gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v10.47765gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v10.47765gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v10.47765gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 3d3a6afd0.47765gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored 3d3a6afd0.47765gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 5f2c057c0.47765gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored 5f2c057c0.47765gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 77adc97b0.47765gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored 77adc97b0.47765gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored cd35dc1f0.47765gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v1 · rescored cd35dc1f0.47765gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored cd35dc1f0.47765gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v10.63165gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v1 · rescored cd35dc1f0.63165gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v10.75465gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v10.75465gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c0.75465gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 5f2c057c0.75465gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 77adc97b0.75465gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 77adc97b0.75465gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f0.75465gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v1 · rescored cd35dc1f0.75465gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored 3d3a6afd0.76965gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored 5f2c057c0.76965gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored 77adc97b0.76965gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored cd35dc1f0.76965gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v20.78565gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v10.78565gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v2 · rescored 3d3a6afd0.78565gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v2 · rescored 5f2c057c0.78565gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v1 · rescored cd35dc1f0.78565gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 3d3a6afd, repeats ×30.81565gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 5f2c057c0.81565gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 5f2c057c, repeats ×30.81565gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 77adc97b0.81565gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 77adc97b, repeats ×30.81565gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored cd35dc1f0.81565gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored cd35dc1f, repeats ×30.81565gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v20.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd0.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c0.86265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v10.89265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v10.89265gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v20.89265gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 3d3a6afd0.89265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored 3d3a6afd0.89265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored 3d3a6afd0.89265gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd0.89265gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 5f2c057c0.89265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored 5f2c057c0.89265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored 5f2c057c0.89265gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c0.89265gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 77adc97b0.89265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored 77adc97b0.89265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored 77adc97b0.89265gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored cd35dc1f0.89265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored cd35dc1f0.89265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored cd35dc1f0.89265gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v20.90865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v10.90865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v20.90865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · repeats0.90865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd0.90865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd0.90865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 5f2c057c0.90865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c0.90865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 5f2c057c0.90865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c, repeats0.90865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b0.90865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b, repeats0.90865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f0.90865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f, repeats0.90865gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 3d3a6afd, repeats ×30.91865gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 5f2c057c, repeats ×30.91865gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 77adc97b, repeats ×30.91865gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored cd35dc1f, repeats ×30.91865gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v20.92365gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd0.92365gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 5f2c057c0.92365gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 5f2c057c, repeats ×30.93365gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 77adc97b, repeats ×30.93365gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored cd35dc1f, repeats ×30.93365gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v20.93865gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v2 · rescored 3d3a6afd0.93865gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v2 · rescored 5f2c057c0.93865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v20.95465gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v10.95465gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v10.95465gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored 3d3a6afd0.95465gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd0.95465gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored 5f2c057c0.95465gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v1 · rescored 5f2c057c0.95465gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c0.95465gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 5f2c057c0.95465gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 5f2c057c0.95465gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored 77adc97b0.95465gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v1 · rescored 77adc97b0.95465gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 77adc97b0.95465gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 77adc97b0.95465gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored cd35dc1f0.95465gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v1 · rescored cd35dc1f0.95465gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored cd35dc1f0.95465gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v1 · rescored cd35dc1f0.95465gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v10.96965gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 3d3a6afd0.96965gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored 3d3a6afd0.96965gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 5f2c057c0.96965gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored 5f2c057c0.96965gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 77adc97b0.96965gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored 77adc97b0.96965gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored cd35dc1f0.96965gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored cd35dc1f0.96965gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v20.98565gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v2 · rescored 3d3a6afd0.98565gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v2 · rescored 5f2c057c0.98565gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v21.03165gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v2 · rescored 3d3a6afd1.03165gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v2 · rescored 5f2c057c1.03165gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v21.04665gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v2 · rescored 3d3a6afd1.04665gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v2 · rescored 5f2c057c1.04665gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v11.06265gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v11.06265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v11.06265gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v1 · rescored 5f2c057c1.06265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v1 · rescored 5f2c057c1.06265gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v1 · rescored 5f2c057c1.06265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v1 · rescored 5f2c057c1.06265gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v1 · rescored 77adc97b1.06265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v1 · rescored 77adc97b1.06265gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v1 · rescored 77adc97b1.06265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v1 · rescored 77adc97b1.06265gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v1 · rescored cd35dc1f1.06265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v1 · rescored cd35dc1f1.06265gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v1 · rescored cd35dc1f1.06265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v1 · rescored cd35dc1f1.06265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v11.07765gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v11.07765gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v1 · rescored 5f2c057c1.07765gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v1 · rescored 5f2c057c1.07765gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v1 · rescored 77adc97b1.07765gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v1 · rescored 77adc97b1.07765gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v1 · rescored cd35dc1f1.07765gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v1 · rescored cd35dc1f1.07765gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v21.09265gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v2 · rescored 3d3a6afd1.09265gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored 3d3a6afd1.09265gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v2 · rescored 5f2c057c1.09265gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored 5f2c057c1.09265gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored 77adc97b1.09265gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored cd35dc1f1.09265gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v21.10865gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v2 · rescored 3d3a6afd1.10865gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v2 · rescored 5f2c057c1.10865gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v11.13865gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored 3d3a6afd1.13865gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored 5f2c057c1.13865gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored 77adc97b1.13865gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored cd35dc1f1.13865gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v11.15465gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored 3d3a6afd1.15465gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored 5f2c057c1.15465gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored 77adc97b1.15465gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored cd35dc1f1.15465gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v11.16965gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v21.16965gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored 3d3a6afd1.16965gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v2 · rescored 3d3a6afd1.16965gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored 5f2c057c1.16965gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v2 · rescored 5f2c057c1.16965gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored 77adc97b1.16965gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored cd35dc1f1.16965gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v21.18565gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v21.18565gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v2 · rescored 3d3a6afd1.18565gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v2 · rescored 3d3a6afd1.18565gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v1 · rescored 5f2c057c1.18565gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v2 · rescored 5f2c057c1.18565gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v2 · rescored 5f2c057c1.18565gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v1 · rescored 77adc97b1.18565gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v1 · rescored cd35dc1f1.18565gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 3d3a6afd, repeats ×31.20065gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 5f2c057c, repeats ×31.20065gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 77adc97b, repeats ×31.20065gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored cd35dc1f, repeats ×31.20065gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v11.21565gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v1 · rescored 5f2c057c1.21565gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c1.21565gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v1 · rescored 77adc97b1.21565gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 77adc97b1.21565gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v1 · rescored cd35dc1f1.21565gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f1.21565gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 3d3a6afd1.24665gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 5f2c057c1.24665gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 77adc97b1.24665gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored cd35dc1f1.24665gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v11.26265gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v21.26265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored 3d3a6afd1.26265gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v2 · rescored 3d3a6afd1.26265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored 5f2c057c1.26265gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v1 · rescored 5f2c057c1.26265gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v2 · rescored 5f2c057c1.26265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored 77adc97b1.26265gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v1 · rescored 77adc97b1.26265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored cd35dc1f1.26265gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v1 · rescored cd35dc1f1.26265gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v21.27765gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v2 · rescored 3d3a6afd1.27765gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v2 · rescored 5f2c057c1.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v21.33865gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v2 · rescored 3d3a6afd1.33865gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v2 · rescored 5f2c057c1.33865gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v11.35465gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored 3d3a6afd1.35465gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored 5f2c057c1.35465gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored 77adc97b1.35465gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored cd35dc1f1.35465gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v11.36965gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v11.36965gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v1 · rescored 5f2c057c1.36965gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v1 · rescored 77adc97b1.36965gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v1 · rescored cd35dc1f1.36965gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored cd35dc1f1.36965gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v11.40065gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored 3d3a6afd1.40065gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored 5f2c057c1.40065gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored 77adc97b1.40065gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored cd35dc1f1.40065gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v11.44665gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v11.44665gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored 3d3a6afd1.44665gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored 3d3a6afd1.44665gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored 5f2c057c1.44665gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored 5f2c057c1.44665gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored 77adc97b1.44665gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored 77adc97b1.44665gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored cd35dc1f1.44665gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored cd35dc1f1.44665gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v21.49265gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored 3d3a6afd1.49265gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v2 · rescored 3d3a6afd1.49265gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored 5f2c057c1.49265gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v2 · rescored 5f2c057c1.49265gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored 77adc97b1.49265gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored cd35dc1f1.49265gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v21.52365gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v11.52365gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v2 · rescored 3d3a6afd1.52365gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored 3d3a6afd1.52365gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v2 · rescored 5f2c057c1.52365gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored 5f2c057c1.52365gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v1 · rescored 5f2c057c1.52365gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored 77adc97b1.52365gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v1 · rescored 77adc97b1.52365gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v1 · rescored cd35dc1f1.52365gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v11.58565gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored 3d3a6afd1.58565gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored 5f2c057c1.58565gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored 77adc97b1.58565gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored cd35dc1f1.58565gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v11.69265gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored 3d3a6afd1.69265gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored 5f2c057c1.69265gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored 77adc97b1.69265gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored cd35dc1f1.69265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v1 · rescored 5f2c057c1.70865gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v1 · rescored 77adc97b1.70865gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v11.81565gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v1 · rescored 5f2c057c1.81565gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v1 · rescored 77adc97b1.81565gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v1 · rescored cd35dc1f1.81565gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v1 · rescored 5f2c057c1.84665gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v1 · rescored 77adc97b1.84665gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v1 · rescored cd35dc1f1.84665gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v1 · rescored 5f2c057c1.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v1 · rescored 5f2c057c1.86265gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v1 · rescored 77adc97b1.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v1 · rescored 77adc97b1.86265gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v11.98565gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v21.98565gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v21.98565gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v2 · rescored 3d3a6afd1.98565gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v2 · rescored 3d3a6afd1.98565gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v1 · rescored 5f2c057c1.98565gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v2 · rescored 5f2c057c1.98565gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v2 · rescored 5f2c057c1.98565gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v1 · rescored 77adc97b1.98565gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v1 · rescored cd35dc1f1.98565gold/orrery_triage_v1_labelled gold
triage_baseline_pro_noreason_v12.07765gold/orrery_triage_v1_labelled gold
triage_baseline_pro_noreason_v1 · rescored 3d3a6afd2.07765gold/orrery_triage_v1_labelled gold
triage_baseline_pro_noreason_v1 · rescored 5f2c057c2.07765gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v12.10865gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored 3d3a6afd2.10865gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored 5f2c057c2.10865gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored 77adc97b2.10865gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored cd35dc1f2.10865gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v22.29265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v2 · rescored 3d3a6afd2.29265gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v2 · rescored 5f2c057c2.29265gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v22.43165gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v2 · rescored 3d3a6afd2.43165gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v2 · rescored 5f2c057c2.43165gold/orrery_triage_v1_labelled gold