eval-harness

metric

n_probes

better
higher
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

How many probes the model asked for before it decided.

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

02on orrery_triage_v1_labelled

gold/orrery_triage_v1_labelledtriage_local_dgx_nemotron_35_lightning_30b_v2: 3triage_local_mac_nemotron_35_lightning_30b_v2: 3triage_local_x3_nemotron_35_lightning_30b_v2: 3triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd: 3triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd: 3triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd: 3triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 5f2c057c: 3triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 5f2c057c: 3triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 5f2c057c: 3triage_local_x3_ministral_3_14b_v1: 2.938triage_local_x3_ministral_3_14b_v1 · rescored 3d3a6afd: 2.938triage_local_x3_ministral_3_14b_v1 · rescored 5f2c057c: 2.938triage_local_x3_ministral_3_14b_v1 · rescored 77adc97b: 2.938triage_local_x3_ministral_3_14b_v1 · rescored cd35dc1f: 2.938triage_hosted_mistral_medium_35_v1: 2.877triage_local_dgx_ministral_3_14b_v1: 2.877triage_local_dgx_phi4_14b_v1: 2.877triage_hosted_mistral_medium_35_v1 · rescored 3d3a6afd: 2.877triage_local_dgx_ministral_3_14b_v1 · rescored 3d3a6afd: 2.877triage_local_dgx_phi4_14b_v1 · rescored 3d3a6afd: 2.877triage_hosted_mistral_medium_35_v1 · rescored 5f2c057c: 2.877triage_local_dgx_ministral_3_14b_v1 · rescored 5f2c057c: 2.877triage_local_dgx_phi4_14b_v1 · rescored 5f2c057c: 2.877triage_hosted_mistral_medium_35_v1 · rescored 77adc97b: 2.877triage_local_dgx_ministral_3_14b_v1 · rescored 77adc97b: 2.877triage_local_dgx_phi4_14b_v1 · rescored 77adc97b: 2.877triage_hosted_mistral_medium_35_v1 · rescored cd35dc1f: 2.877triage_local_dgx_ministral_3_14b_v1 · rescored cd35dc1f: 2.877triage_local_dgx_phi4_14b_v1 · rescored cd35dc1f: 2.877triage_local_dgx_qwen3_30b_a3b_v1: 2.862triage_local_dgx_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 2.862triage_local_dgx_qwen3_30b_a3b_v1 · rescored 5f2c057c: 2.862triage_local_dgx_qwen3_30b_a3b_v1 · rescored 77adc97b: 2.862triage_local_dgx_qwen3_30b_a3b_v1 · rescored cd35dc1f: 2.862triage_local_mac_ministral_3_14b_v1: 2.831triage_local_mac_qwen3_30b_a3b_v1: 2.831triage_local_x3_phi4_14b_v1: 2.831triage_local_mac_ministral_3_14b_v1 · rescored 3d3a6afd: 2.831triage_local_mac_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 2.831triage_local_x3_phi4_14b_v1 · rescored 3d3a6afd: 2.831triage_local_mac_ministral_3_14b_v1 · rescored 5f2c057c: 2.831triage_local_mac_qwen3_30b_a3b_v1 · rescored 5f2c057c: 2.831triage_local_x3_phi4_14b_v1 · rescored 5f2c057c: 2.831triage_local_mac_ministral_3_14b_v1 · rescored 77adc97b: 2.831triage_local_mac_qwen3_30b_a3b_v1 · rescored 77adc97b: 2.831triage_local_x3_phi4_14b_v1 · rescored 77adc97b: 2.831triage_local_mac_ministral_3_14b_v1 · rescored cd35dc1f: 2.831triage_local_mac_qwen3_30b_a3b_v1 · rescored cd35dc1f: 2.831triage_local_x3_phi4_14b_v1 · rescored cd35dc1f: 2.831triage_local_mac_llama31_8b_v1: 2.8triage_local_mac_phi4_14b_v1: 2.8triage_local_x3_qwen3_30b_a3b_v1: 2.8triage_local_mac_llama31_8b_v1 · rescored 3d3a6afd: 2.8triage_local_mac_phi4_14b_v1 · rescored 3d3a6afd: 2.8triage_local_x3_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 2.8triage_local_mac_llama31_8b_v1 · rescored 5f2c057c: 2.8triage_local_mac_phi4_14b_v1 · rescored 5f2c057c: 2.8triage_local_x3_qwen3_30b_a3b_v1 · rescored 5f2c057c: 2.8triage_local_mac_llama31_8b_v1 · rescored 77adc97b: 2.8triage_local_mac_phi4_14b_v1 · rescored 77adc97b: 2.8triage_local_x3_qwen3_30b_a3b_v1 · rescored 77adc97b: 2.8triage_local_mac_llama31_8b_v1 · rescored cd35dc1f: 2.8triage_local_mac_phi4_14b_v1 · rescored cd35dc1f: 2.8triage_local_x3_qwen3_30b_a3b_v1 · rescored cd35dc1f: 2.8triage_local_dgx_llama31_8b_v1: 2.785triage_local_dgx_llama31_8b_v1 · rescored 3d3a6afd: 2.785triage_local_dgx_llama31_8b_v1 · rescored 5f2c057c: 2.785triage_local_dgx_llama31_8b_v1 · rescored 77adc97b: 2.785triage_local_dgx_llama31_8b_v1 · rescored cd35dc1f: 2.785triage_local_x3_llama31_8b_v1: 2.769triage_local_x3_llama31_8b_v1 · rescored 3d3a6afd: 2.769triage_local_x3_llama31_8b_v1 · rescored 5f2c057c: 2.769triage_local_x3_llama31_8b_v1 · rescored 77adc97b: 2.769triage_local_x3_llama31_8b_v1 · rescored cd35dc1f: 2.769triage_local_dgx_mistral_medium_35_128b_v1: 2.738triage_local_dgx_mistral_medium_35_128b_v1 · rescored 3d3a6afd: 2.738triage_local_dgx_mistral_medium_35_128b_v1 · rescored 5f2c057c: 2.738triage_local_dgx_mistral_medium_35_128b_v1 · rescored 77adc97b: 2.738triage_local_dgx_mistral_medium_35_128b_v1 · rescored cd35dc1f: 2.738triage_hosted_qwen3_30b_a3b_v1: 2.723triage_hosted_qwen3_30b_a3b_v1 · rescored 3d3a6afd: 2.723triage_hosted_qwen3_30b_a3b_v1 · rescored 5f2c057c: 2.723triage_hosted_qwen3_30b_a3b_v1 · rescored 77adc97b: 2.723triage_hosted_qwen3_30b_a3b_v1 · rescored cd35dc1f: 2.723triage_local_mac_nemotron_35_bolt_e4b_v2: 2.677triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd: 2.677triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c: 2.677triage_local_dgx_nemotron_35_bolt_e4b_v2: 2.6triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd: 2.6triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c: 2.6triage_mistral_large_4_v1: 2.554triage_mistral_large_4_v1 · rescored 5f2c057c: 2.554triage_mistral_large_4_v1 · rescored 77adc97b: 2.554triage_mistral_large_4_v1 · rescored cd35dc1f: 2.554triage_local_dgx_llama4_scout_v1: 2.538triage_mistral_large_4_v2: 2.538triage_local_dgx_llama4_scout_v1 · rescored 3d3a6afd: 2.538triage_mistral_large_4_v2 · rescored 3d3a6afd: 2.538triage_local_dgx_llama4_scout_v1 · rescored 5f2c057c: 2.538triage_mistral_large_4_v2 · rescored 5f2c057c: 2.538triage_local_dgx_llama4_scout_v1 · rescored 77adc97b: 2.538triage_local_dgx_llama4_scout_v1 · rescored cd35dc1f: 2.538triage_baseline_pro_noreason_v1: 2.508triage_baseline_pro_noreason_v1 · rescored 3d3a6afd: 2.508triage_baseline_pro_noreason_v1 · rescored 5f2c057c: 2.508triage_local_x3_nemotron_35_bolt_e4b_v2: 2.492triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd: 2.492triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c: 2.492triage_local_mac_nemotron_35_bolt_e4b_v1: 2.4triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c: 2.4triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 77adc97b: 2.4triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f: 2.4triage_local_dgx_nemotron_35_bolt_e4b_v1: 2.385triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c: 2.385triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 77adc97b: 2.385triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f: 2.385triage_local_dgx_nemotron_35_lightning_30b_v1: 2.338triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 5f2c057c: 2.338triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 77adc97b: 2.338triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored cd35dc1f: 2.338triage_local_x3_nemotron_35_bolt_e4b_v1: 2.323triage_local_x3_nemotron_35_bolt_e4b_v1 · repeats: 2.323triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c: 2.323triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c, repeats: 2.323triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b: 2.323triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b, repeats: 2.323triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f: 2.323triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f, repeats: 2.323triage_local_dgx_nemotron_35_super_120b_v2: 2.308triage_local_dgx_nemotron_35_super_120b_v2 · rescored 3d3a6afd: 2.308triage_local_dgx_nemotron_35_super_120b_v2 · rescored 5f2c057c: 2.308triage_gemini_4_argon_v1: 2.292triage_gemini_4_argon_v1 · rescored 3d3a6afd: 2.292triage_gemini_4_argon_v1 · rescored 5f2c057c: 2.292triage_gemini_4_argon_v1 · rescored 77adc97b: 2.292triage_gemini_4_argon_v1 · rescored cd35dc1f: 2.292triage_local_mac_nemotron_35_lightning_30b_v1: 2.277triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 5f2c057c: 2.277triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 77adc97b: 2.277triage_local_mac_nemotron_35_lightning_30b_v1 · rescored cd35dc1f: 2.277triage_local_x3_nemotron_35_lightning_30b_v1: 2.246triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 5f2c057c: 2.246triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 77adc97b: 2.246triage_local_x3_nemotron_35_lightning_30b_v1 · rescored cd35dc1f: 2.246triage_deepseek_41flash_v2: 2.138triage_deepseek_41flash_v2 · rescored 3d3a6afd: 2.138triage_deepseek_41flash_v2 · rescored 5f2c057c: 2.138triage_hosted_qwen35_122b_v1: 2.123triage_hosted_qwen35_122b_v1 · rescored 5f2c057c: 2.123triage_hosted_qwen35_122b_v1 · rescored 77adc97b: 2.123triage_hosted_qwen35_122b_v1 · rescored cd35dc1f: 2.123triage_gpt_6_luna_v1: 2.108triage_gpt_6_luna_v1 · rescored 5f2c057c: 2.108triage_gpt_6_luna_v1 · rescored 77adc97b: 2.108triage_gpt_6_luna_v1 · rescored cd35dc1f: 2.108triage_kimi_k31_pro_v1: 2.062triage_kimi_k31_pro_v1 · rescored 5f2c057c: 2.062triage_kimi_k31_pro_v1 · rescored 77adc97b: 2.062triage_kimi_k31_pro_v1 · rescored cd35dc1f: 2.062triage_gpt_6_luna_v1 · repeats ×3: 2.041triage_gpt_6_luna_v1 · rescored 5f2c057c, repeats ×3: 2.041triage_gpt_6_luna_v1 · rescored 77adc97b, repeats ×3: 2.041triage_gpt_6_luna_v1 · rescored cd35dc1f, repeats ×3: 2.041triage_local_dgx_qwen35_122b_v1: 2.031triage_local_dgx_qwen35_122b_v1 · rescored 5f2c057c: 2.031triage_local_dgx_qwen35_122b_v1 · rescored 77adc97b: 2.031triage_local_dgx_qwen35_122b_v1 · rescored cd35dc1f: 2.031triage_kimi_k31_pro_v2: 2.015triage_kimi_k31_pro_v2 · rescored 3d3a6afd: 2.015triage_kimi_k31_pro_v2 · rescored 5f2c057c: 2.015triage_local_dgx_qwen38_27b_v2: 1.985triage_local_dgx_qwen38_27b_v2 · rescored 3d3a6afd: 1.985triage_local_dgx_qwen38_27b_v2 · rescored 5f2c057c: 1.985triage_local_mac_qwen38_27b_v2: 1.969triage_local_mac_qwen38_27b_v2 · rescored 3d3a6afd: 1.969triage_local_mac_qwen38_27b_v2 · rescored 5f2c057c: 1.969triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1: 1.954triage_local_x3_qwen38_27b_v2: 1.954triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 3d3a6afd: 1.954triage_local_x3_qwen38_27b_v2 · rescored 3d3a6afd: 1.954triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 5f2c057c: 1.954triage_local_x3_qwen38_27b_v2 · rescored 5f2c057c: 1.954triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 77adc97b: 1.954triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored cd35dc1f: 1.954triage_local_mac_gemma4_31b_v2: 1.908triage_local_mac_gemma4_31b_v2 · rescored 3d3a6afd: 1.908triage_local_mac_gemma4_31b_v2 · rescored 5f2c057c: 1.908triage_hosted_gemma4_31b_v1: 1.892triage_local_dgx_gemma4_31b_v2: 1.892triage_local_x3_gemma4_31b_v1: 1.892triage_local_dgx_gemma4_31b_v2 · rescored 3d3a6afd: 1.892triage_hosted_gemma4_31b_v1 · rescored 5f2c057c: 1.892triage_local_dgx_gemma4_31b_v2 · rescored 5f2c057c: 1.892triage_local_x3_gemma4_31b_v1 · rescored 5f2c057c: 1.892triage_hosted_gemma4_31b_v1 · rescored 77adc97b: 1.892triage_local_x3_gemma4_31b_v1 · rescored 77adc97b: 1.892triage_hosted_gemma4_31b_v1 · rescored cd35dc1f: 1.892triage_local_x3_gemma4_31b_v1 · rescored cd35dc1f: 1.892triage_hosted_gemma4_31b_v2: 1.862triage_local_dgx_gemma4_31b_v1: 1.862triage_local_mac_gemma4_31b_v1: 1.862triage_hosted_gemma4_31b_v2 · rescored 3d3a6afd: 1.862triage_hosted_gemma4_31b_v2 · rescored 5f2c057c: 1.862triage_local_dgx_gemma4_31b_v1 · rescored 5f2c057c: 1.862triage_local_mac_gemma4_31b_v1 · rescored 5f2c057c: 1.862triage_local_dgx_gemma4_31b_v1 · rescored 77adc97b: 1.862triage_local_mac_gemma4_31b_v1 · rescored 77adc97b: 1.862triage_local_dgx_gemma4_31b_v1 · rescored cd35dc1f: 1.862triage_local_mac_gemma4_31b_v1 · rescored cd35dc1f: 1.862triage_local_x3_gemma4_31b_v2: 1.846triage_local_x3_gemma4_31b_v2 · rescored 3d3a6afd: 1.846triage_local_x3_gemma4_31b_v2 · rescored 5f2c057c: 1.846triage_gpt_61_sol_v1: 1.831triage_gpt_61_sol_v1 · rescored 3d3a6afd: 1.831triage_gpt_61_sol_v1 · rescored 5f2c057c: 1.831triage_gpt_61_sol_v1 · rescored 77adc97b: 1.831triage_gpt_61_sol_v1 · rescored cd35dc1f: 1.831triage_gpt_61_sol_v1 · repeats ×3: 1.81triage_gpt_61_sol_v1 · rescored 3d3a6afd, repeats ×3: 1.81triage_gpt_61_sol_v1 · rescored 5f2c057c, repeats ×3: 1.81triage_gpt_61_sol_v1 · rescored 77adc97b, repeats ×3: 1.81triage_gpt_61_sol_v1 · rescored cd35dc1f, repeats ×3: 1.81triage_hosted_qwen35_122b_v2: 1.8triage_hosted_qwen35_122b_v2 · rescored 3d3a6afd: 1.8triage_hosted_qwen35_122b_v2 · rescored 5f2c057c: 1.8triage_local_dgx_qwen35_122b_v2: 1.754triage_sonnet_55_v1: 1.754triage_local_dgx_qwen35_122b_v2 · rescored 3d3a6afd: 1.754triage_sonnet_55_v1 · rescored 3d3a6afd: 1.754triage_local_dgx_qwen35_122b_v2 · rescored 5f2c057c: 1.754triage_sonnet_55_v1 · rescored 5f2c057c: 1.754triage_sonnet_55_v1 · rescored 77adc97b: 1.754triage_sonnet_55_v1 · rescored cd35dc1f: 1.754triage_deepseek_41flash_v1: 1.738triage_deepseek_41flash_v1 · rescored 5f2c057c: 1.738triage_deepseek_41flash_v1 · rescored 77adc97b: 1.738triage_deepseek_41flash_v1 · rescored cd35dc1f: 1.738triage_baseline_pro_v1 · repeats ×3: 1.6triage_baseline_pro_v1 · rescored 3d3a6afd, repeats ×3: 1.6triage_baseline_pro_v1 · rescored 5f2c057c, repeats ×3: 1.6triage_baseline_pro_v1 · rescored 77adc97b, repeats ×3: 1.6triage_baseline_pro_v1 · rescored cd35dc1f, repeats ×3: 1.6triage_local_dgx_qwen38_27b_v1: 1.569triage_local_dgx_qwen38_27b_v1 · rescored 5f2c057c: 1.569triage_local_dgx_qwen38_27b_v1 · rescored 77adc97b: 1.569triage_local_dgx_qwen38_27b_v1 · rescored cd35dc1f: 1.569triage_baseline_pro_v1: 1.538triage_baseline_pro_v1 · rescored 3d3a6afd: 1.538triage_baseline_pro_v1 · rescored 5f2c057c: 1.538triage_baseline_pro_v1 · rescored 77adc97b: 1.538triage_baseline_pro_v1 · rescored cd35dc1f: 1.538triage_local_mac_qwen38_27b_v1: 1.523triage_local_mac_qwen38_27b_v1 · rescored 5f2c057c: 1.523triage_local_mac_qwen38_27b_v1 · rescored 77adc97b: 1.523triage_local_mac_qwen38_27b_v1 · rescored cd35dc1f: 1.523triage_local_x3_qwen38_27b_v1: 1.477triage_local_x3_qwen38_27b_v1 · rescored 5f2c057c: 1.477triage_local_x3_qwen38_27b_v1 · rescored 77adc97b: 1.477triage_local_x3_qwen38_27b_v1 · rescored cd35dc1f: 1.477triage_local_dgx_nemotron_35_super_120b_v1: 1.446triage_local_dgx_nemotron_35_super_120b_v1 · rescored 5f2c057c: 1.446triage_local_dgx_nemotron_35_super_120b_v1 · rescored 77adc97b: 1.446triage_local_dgx_nemotron_35_super_120b_v1 · rescored cd35dc1f: 1.446triage_gpt_6_luna_v2: 1.415triage_gpt_6_luna_v2 · rescored 3d3a6afd: 1.415triage_gpt_6_luna_v2 · rescored 5f2c057c: 1.415triage_hosted_gpt_oss_120b_v1: 1.338triage_hosted_gpt_oss_120b_v1 · rescored 5f2c057c: 1.338triage_hosted_gpt_oss_120b_v1 · rescored 77adc97b: 1.338triage_hosted_gpt_oss_120b_v1 · rescored cd35dc1f: 1.338triage_haiku_55_v1 · repeats ×3: 1.333triage_haiku_55_v1 · rescored 3d3a6afd, repeats ×3: 1.333triage_haiku_55_v1 · rescored 5f2c057c, repeats ×3: 1.333triage_haiku_55_v1 · rescored 77adc97b, repeats ×3: 1.333triage_haiku_55_v1 · rescored cd35dc1f, repeats ×3: 1.333triage_local_dgx_gemma4_12b_noreason_v1: 1.323triage_local_dgx_gemma4_12b_noreason_v1 · rescored 3d3a6afd: 1.323triage_local_dgx_gemma4_12b_noreason_v1 · rescored 5f2c057c: 1.323triage_local_dgx_gemma4_12b_noreason_v1 · rescored 77adc97b: 1.323triage_local_dgx_gemma4_12b_noreason_v1 · rescored cd35dc1f: 1.323triage_haiku_55_v1: 1.277triage_hosted_gpt_oss_120b_v2: 1.277triage_local_mac_gemma4_12b_noreason_v1: 1.277triage_local_mac_gpt_oss_20b_v2: 1.277triage_haiku_55_v1 · rescored 3d3a6afd: 1.277triage_hosted_gpt_oss_120b_v2 · rescored 3d3a6afd: 1.277triage_local_mac_gemma4_12b_noreason_v1 · rescored 3d3a6afd: 1.277triage_local_mac_gpt_oss_20b_v2 · rescored 3d3a6afd: 1.277triage_haiku_55_v1 · rescored 5f2c057c: 1.277triage_hosted_gpt_oss_120b_v2 · rescored 5f2c057c: 1.277triage_local_mac_gemma4_12b_noreason_v1 · rescored 5f2c057c: 1.277triage_local_mac_gpt_oss_20b_v2 · rescored 5f2c057c: 1.277triage_haiku_55_v1 · rescored 77adc97b: 1.277triage_local_mac_gemma4_12b_noreason_v1 · rescored 77adc97b: 1.277triage_haiku_55_v1 · rescored cd35dc1f: 1.277triage_local_mac_gemma4_12b_noreason_v1 · rescored cd35dc1f: 1.277triage_local_x3_gpt_oss_20b_v2: 1.262triage_local_x3_gpt_oss_20b_v2 · rescored 3d3a6afd: 1.262triage_local_x3_gpt_oss_20b_v2 · rescored 5f2c057c: 1.262triage_gemini_35_flash_lite_v1: 1.2triage_gemini_35_flash_lite_v1 · rescored 3d3a6afd: 1.2triage_gemini_35_flash_lite_v1 · rescored 5f2c057c: 1.2triage_gemini_35_flash_lite_v1 · rescored 77adc97b: 1.2triage_gemini_35_flash_lite_v1 · rescored cd35dc1f: 1.2triage_local_dgx_gpt_oss_120b_v2: 1.185triage_local_dgx_gpt_oss_120b_v2 · rescored 3d3a6afd: 1.185triage_local_dgx_gpt_oss_120b_v2 · rescored 5f2c057c: 1.185triage_local_dgx_gpt_oss_20b_v2: 1.123triage_local_dgx_gpt_oss_20b_v2 · rescored 3d3a6afd: 1.123triage_local_dgx_gpt_oss_20b_v2 · rescored 5f2c057c: 1.123triage_local_dgx_gpt_oss_120b_v1: 1.031triage_local_dgx_gpt_oss_120b_v1 · rescored 5f2c057c: 1.031triage_local_dgx_gpt_oss_120b_v1 · rescored 77adc97b: 1.031triage_local_dgx_gpt_oss_120b_v1 · rescored cd35dc1f: 1.031triage_stub_dismiss_v1: 1triage_stub_escalate_v1: 1triage_stub_dismiss_v1 · rescored 3d3a6afd: 1triage_stub_escalate_v1 · rescored 3d3a6afd: 1triage_stub_dismiss_v1 · rescored 5f2c057c: 1triage_stub_escalate_v1 · rescored 5f2c057c: 1triage_stub_dismiss_v1 · rescored 77adc97b: 1triage_stub_escalate_v1 · rescored 77adc97b: 1triage_stub_dismiss_v1 · rescored cd35dc1f: 1triage_stub_escalate_v1 · rescored cd35dc1f: 1triage_local_dgx_gpt_oss_20b_v1: 0.785triage_local_dgx_gpt_oss_20b_v1 · rescored 5f2c057c: 0.785triage_local_dgx_gpt_oss_20b_v1 · rescored 77adc97b: 0.785triage_local_dgx_gpt_oss_20b_v1 · rescored cd35dc1f: 0.785triage_local_x3_gpt_oss_20b_v1: 0.738triage_local_x3_gpt_oss_20b_v1 · rescored 5f2c057c: 0.738triage_local_x3_gpt_oss_20b_v1 · rescored 77adc97b: 0.738triage_local_x3_gpt_oss_20b_v1 · rescored cd35dc1f: 0.738triage_local_mac_gpt_oss_20b_v1: 0.723triage_local_mac_gpt_oss_20b_v1 · rescored 5f2c057c: 0.723triage_local_mac_gpt_oss_20b_v1 · rescored 77adc97b: 0.723triage_local_mac_gpt_oss_20b_v1 · rescored cd35dc1f: 0.7230.723better · 3

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 n_probesitemsscored against
triage_local_dgx_nemotron_35_lightning_30b_v23.00065gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v23.00065gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v23.00065gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd3.00065gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd3.00065gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 3d3a6afd3.00065gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v2 · rescored 5f2c057c3.00065gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v2 · rescored 5f2c057c3.00065gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v2 · rescored 5f2c057c3.00065gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v12.93865gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored 3d3a6afd2.93865gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored 5f2c057c2.93865gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored 77adc97b2.93865gold/orrery_triage_v1_labelled gold
triage_local_x3_ministral_3_14b_v1 · rescored cd35dc1f2.93865gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v12.87765gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v12.87765gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v12.87765gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored 3d3a6afd2.87765gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored 3d3a6afd2.87765gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored 3d3a6afd2.87765gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored 5f2c057c2.87765gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored 5f2c057c2.87765gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored 5f2c057c2.87765gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored 77adc97b2.87765gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored 77adc97b2.87765gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored 77adc97b2.87765gold/orrery_triage_v1_labelled gold
triage_hosted_mistral_medium_35_v1 · rescored cd35dc1f2.87765gold/orrery_triage_v1_labelled gold
triage_local_dgx_ministral_3_14b_v1 · rescored cd35dc1f2.87765gold/orrery_triage_v1_labelled gold
triage_local_dgx_phi4_14b_v1 · rescored cd35dc1f2.87765gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v12.86265gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored 3d3a6afd2.86265gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored 5f2c057c2.86265gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored 77adc97b2.86265gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen3_30b_a3b_v1 · rescored cd35dc1f2.86265gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v12.83165gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v12.83165gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v12.83165gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored 3d3a6afd2.83165gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored 3d3a6afd2.83165gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored 3d3a6afd2.83165gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored 5f2c057c2.83165gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored 5f2c057c2.83165gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored 5f2c057c2.83165gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored 77adc97b2.83165gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored 77adc97b2.83165gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored 77adc97b2.83165gold/orrery_triage_v1_labelled gold
triage_local_mac_ministral_3_14b_v1 · rescored cd35dc1f2.83165gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen3_30b_a3b_v1 · rescored cd35dc1f2.83165gold/orrery_triage_v1_labelled gold
triage_local_x3_phi4_14b_v1 · rescored cd35dc1f2.83165gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v12.80065gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v12.80065gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v12.80065gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored 3d3a6afd2.80065gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored 3d3a6afd2.80065gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored 3d3a6afd2.80065gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored 5f2c057c2.80065gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored 5f2c057c2.80065gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored 5f2c057c2.80065gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored 77adc97b2.80065gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored 77adc97b2.80065gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored 77adc97b2.80065gold/orrery_triage_v1_labelled gold
triage_local_mac_llama31_8b_v1 · rescored cd35dc1f2.80065gold/orrery_triage_v1_labelled gold
triage_local_mac_phi4_14b_v1 · rescored cd35dc1f2.80065gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen3_30b_a3b_v1 · rescored cd35dc1f2.80065gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v12.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored 3d3a6afd2.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored 5f2c057c2.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored 77adc97b2.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama31_8b_v1 · rescored cd35dc1f2.78565gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v12.76965gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored 3d3a6afd2.76965gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored 5f2c057c2.76965gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored 77adc97b2.76965gold/orrery_triage_v1_labelled gold
triage_local_x3_llama31_8b_v1 · rescored cd35dc1f2.76965gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v12.73865gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored 3d3a6afd2.73865gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored 5f2c057c2.73865gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored 77adc97b2.73865gold/orrery_triage_v1_labelled gold
triage_local_dgx_mistral_medium_35_128b_v1 · rescored cd35dc1f2.73865gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v12.72365gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored 3d3a6afd2.72365gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored 5f2c057c2.72365gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored 77adc97b2.72365gold/orrery_triage_v1_labelled gold
triage_hosted_qwen3_30b_a3b_v1 · rescored cd35dc1f2.72365gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v22.67765gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd2.67765gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c2.67765gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v22.60065gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd2.60065gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c2.60065gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v12.55465gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v1 · rescored 5f2c057c2.55465gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v1 · rescored 77adc97b2.55465gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v1 · rescored cd35dc1f2.55465gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v12.53865gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v22.53865gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored 3d3a6afd2.53865gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v2 · rescored 3d3a6afd2.53865gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored 5f2c057c2.53865gold/orrery_triage_v1_labelled gold
triage_mistral_large_4_v2 · rescored 5f2c057c2.53865gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored 77adc97b2.53865gold/orrery_triage_v1_labelled gold
triage_local_dgx_llama4_scout_v1 · rescored cd35dc1f2.53865gold/orrery_triage_v1_labelled gold
triage_baseline_pro_noreason_v12.50865gold/orrery_triage_v1_labelled gold
triage_baseline_pro_noreason_v1 · rescored 3d3a6afd2.50865gold/orrery_triage_v1_labelled gold
triage_baseline_pro_noreason_v1 · rescored 5f2c057c2.50865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v22.49265gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 3d3a6afd2.49265gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v2 · rescored 5f2c057c2.49265gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v12.40065gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c2.40065gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored 77adc97b2.40065gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f2.40065gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v12.38565gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c2.38565gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored 77adc97b2.38565gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f2.38565gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v12.33865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 5f2c057c2.33865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored 77adc97b2.33865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_lightning_30b_v1 · rescored cd35dc1f2.33865gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v12.32365gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · repeats2.32365gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c2.32365gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 5f2c057c, repeats2.32365gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b2.32365gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored 77adc97b, repeats2.32365gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f2.32365gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_bolt_e4b_v1 · rescored cd35dc1f, repeats2.32365gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v22.30865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v2 · rescored 3d3a6afd2.30865gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v2 · rescored 5f2c057c2.30865gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v12.29265gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored 3d3a6afd2.29265gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored 5f2c057c2.29265gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored 77adc97b2.29265gold/orrery_triage_v1_labelled gold
triage_gemini_4_argon_v1 · rescored cd35dc1f2.29265gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v12.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 5f2c057c2.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v1 · rescored 77adc97b2.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_nemotron_35_lightning_30b_v1 · rescored cd35dc1f2.27765gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v12.24665gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 5f2c057c2.24665gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v1 · rescored 77adc97b2.24665gold/orrery_triage_v1_labelled gold
triage_local_x3_nemotron_35_lightning_30b_v1 · rescored cd35dc1f2.24665gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v22.13865gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v2 · rescored 3d3a6afd2.13865gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v2 · rescored 5f2c057c2.13865gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v12.12365gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v1 · rescored 5f2c057c2.12365gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v1 · rescored 77adc97b2.12365gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v1 · rescored cd35dc1f2.12365gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v12.10865gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 5f2c057c2.10865gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 77adc97b2.10865gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored cd35dc1f2.10865gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v12.06265gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v1 · rescored 5f2c057c2.06265gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v1 · rescored 77adc97b2.06265gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v1 · rescored cd35dc1f2.06265gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · repeats ×32.04165gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 5f2c057c, repeats ×32.04165gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored 77adc97b, repeats ×32.04165gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v1 · rescored cd35dc1f, repeats ×32.04165gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v12.03165gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v1 · rescored 5f2c057c2.03165gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v1 · rescored 77adc97b2.03165gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v1 · rescored cd35dc1f2.03165gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v22.01565gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v2 · rescored 3d3a6afd2.01565gold/orrery_triage_v1_labelled gold
triage_kimi_k31_pro_v2 · rescored 5f2c057c2.01565gold/orrery_triage_v1_labelled gold
triage_local_dgx_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_dgx_qwen38_27b_v2 · rescored 5f2c057c1.98565gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v21.96965gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v2 · rescored 3d3a6afd1.96965gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v2 · rescored 5f2c057c1.96965gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v11.95465gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v21.95465gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 3d3a6afd1.95465gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v2 · rescored 3d3a6afd1.95465gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 5f2c057c1.95465gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v2 · rescored 5f2c057c1.95465gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 77adc97b1.95465gold/orrery_triage_v1_labelled gold
triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored cd35dc1f1.95465gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v21.90865gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v2 · rescored 3d3a6afd1.90865gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v2 · rescored 5f2c057c1.90865gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v11.89265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v21.89265gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v11.89265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v2 · rescored 3d3a6afd1.89265gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v1 · rescored 5f2c057c1.89265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v2 · rescored 5f2c057c1.89265gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v1 · rescored 5f2c057c1.89265gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v1 · rescored 77adc97b1.89265gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v1 · rescored 77adc97b1.89265gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v1 · rescored cd35dc1f1.89265gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v1 · rescored cd35dc1f1.89265gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v21.86265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v11.86265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v11.86265gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v2 · rescored 3d3a6afd1.86265gold/orrery_triage_v1_labelled gold
triage_hosted_gemma4_31b_v2 · rescored 5f2c057c1.86265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v1 · rescored 5f2c057c1.86265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v1 · rescored 5f2c057c1.86265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v1 · rescored 77adc97b1.86265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v1 · rescored 77adc97b1.86265gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_31b_v1 · rescored cd35dc1f1.86265gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_31b_v1 · rescored cd35dc1f1.86265gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v21.84665gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v2 · rescored 3d3a6afd1.84665gold/orrery_triage_v1_labelled gold
triage_local_x3_gemma4_31b_v2 · rescored 5f2c057c1.84665gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v11.83165gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 3d3a6afd1.83165gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 5f2c057c1.83165gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 77adc97b1.83165gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored cd35dc1f1.83165gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · repeats ×31.81065gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 3d3a6afd, repeats ×31.81065gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 5f2c057c, repeats ×31.81065gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored 77adc97b, repeats ×31.81065gold/orrery_triage_v1_labelled gold
triage_gpt_61_sol_v1 · rescored cd35dc1f, repeats ×31.81065gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v21.80065gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v2 · rescored 3d3a6afd1.80065gold/orrery_triage_v1_labelled gold
triage_hosted_qwen35_122b_v2 · rescored 5f2c057c1.80065gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v21.75465gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v11.75465gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v2 · rescored 3d3a6afd1.75465gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored 3d3a6afd1.75465gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen35_122b_v2 · rescored 5f2c057c1.75465gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored 5f2c057c1.75465gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored 77adc97b1.75465gold/orrery_triage_v1_labelled gold
triage_sonnet_55_v1 · rescored cd35dc1f1.75465gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v11.73865gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v1 · rescored 5f2c057c1.73865gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v1 · rescored 77adc97b1.73865gold/orrery_triage_v1_labelled gold
triage_deepseek_41flash_v1 · rescored cd35dc1f1.73865gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · repeats ×31.60065gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 3d3a6afd, repeats ×31.60065gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 5f2c057c, repeats ×31.60065gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 77adc97b, repeats ×31.60065gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored cd35dc1f, repeats ×31.60065gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v11.56965gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v1 · rescored 5f2c057c1.56965gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v1 · rescored 77adc97b1.56965gold/orrery_triage_v1_labelled gold
triage_local_dgx_qwen38_27b_v1 · rescored cd35dc1f1.56965gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v11.53865gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 3d3a6afd1.53865gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 5f2c057c1.53865gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored 77adc97b1.53865gold/orrery_triage_v1_labelled gold
triage_baseline_pro_v1 · rescored cd35dc1f1.53865gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v11.52365gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v1 · rescored 5f2c057c1.52365gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v1 · rescored 77adc97b1.52365gold/orrery_triage_v1_labelled gold
triage_local_mac_qwen38_27b_v1 · rescored cd35dc1f1.52365gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v11.47765gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v1 · rescored 5f2c057c1.47765gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v1 · rescored 77adc97b1.47765gold/orrery_triage_v1_labelled gold
triage_local_x3_qwen38_27b_v1 · rescored cd35dc1f1.47765gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v11.44665gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v1 · rescored 5f2c057c1.44665gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v1 · rescored 77adc97b1.44665gold/orrery_triage_v1_labelled gold
triage_local_dgx_nemotron_35_super_120b_v1 · rescored cd35dc1f1.44665gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v21.41565gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v2 · rescored 3d3a6afd1.41565gold/orrery_triage_v1_labelled gold
triage_gpt_6_luna_v2 · rescored 5f2c057c1.41565gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v11.33865gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v1 · rescored 5f2c057c1.33865gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v1 · rescored 77adc97b1.33865gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v1 · rescored cd35dc1f1.33865gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · repeats ×31.33365gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 3d3a6afd, repeats ×31.33365gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 5f2c057c, repeats ×31.33365gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 77adc97b, repeats ×31.33365gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored cd35dc1f, repeats ×31.33365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v11.32365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored 3d3a6afd1.32365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored 5f2c057c1.32365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored 77adc97b1.32365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gemma4_12b_noreason_v1 · rescored cd35dc1f1.32365gold/orrery_triage_v1_labelled gold
triage_haiku_55_v11.27765gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v21.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v11.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v21.27765gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 3d3a6afd1.27765gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v2 · rescored 3d3a6afd1.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored 3d3a6afd1.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v2 · rescored 3d3a6afd1.27765gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 5f2c057c1.27765gold/orrery_triage_v1_labelled gold
triage_hosted_gpt_oss_120b_v2 · rescored 5f2c057c1.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored 5f2c057c1.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v2 · rescored 5f2c057c1.27765gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored 77adc97b1.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored 77adc97b1.27765gold/orrery_triage_v1_labelled gold
triage_haiku_55_v1 · rescored cd35dc1f1.27765gold/orrery_triage_v1_labelled gold
triage_local_mac_gemma4_12b_noreason_v1 · rescored cd35dc1f1.27765gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v21.26265gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v2 · rescored 3d3a6afd1.26265gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v2 · rescored 5f2c057c1.26265gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v11.20065gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored 3d3a6afd1.20065gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored 5f2c057c1.20065gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored 77adc97b1.20065gold/orrery_triage_v1_labelled gold
triage_gemini_35_flash_lite_v1 · rescored cd35dc1f1.20065gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v21.18565gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v2 · rescored 3d3a6afd1.18565gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v2 · rescored 5f2c057c1.18565gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v21.12365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v2 · rescored 3d3a6afd1.12365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v2 · rescored 5f2c057c1.12365gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v11.03165gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v1 · rescored 5f2c057c1.03165gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v1 · rescored 77adc97b1.03165gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_120b_v1 · rescored cd35dc1f1.03165gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v11.00065gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v11.00065gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored 3d3a6afd1.00065gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored 3d3a6afd1.00065gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored 5f2c057c1.00065gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored 5f2c057c1.00065gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored 77adc97b1.00065gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored 77adc97b1.00065gold/orrery_triage_v1_labelled gold
triage_stub_dismiss_v1 · rescored cd35dc1f1.00065gold/orrery_triage_v1_labelled gold
triage_stub_escalate_v1 · rescored cd35dc1f1.00065gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v10.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v1 · rescored 5f2c057c0.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v1 · rescored 77adc97b0.78565gold/orrery_triage_v1_labelled gold
triage_local_dgx_gpt_oss_20b_v1 · rescored cd35dc1f0.78565gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v10.73865gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v1 · rescored 5f2c057c0.73865gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v1 · rescored 77adc97b0.73865gold/orrery_triage_v1_labelled gold
triage_local_x3_gpt_oss_20b_v1 · rescored cd35dc1f0.73865gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v10.72365gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v1 · rescored 5f2c057c0.72365gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v1 · rescored 77adc97b0.72365gold/orrery_triage_v1_labelled gold
triage_local_mac_gpt_oss_20b_v1 · rescored cd35dc1f0.72365gold/orrery_triage_v1_labelled gold