capability-summary.json

{
 "note": "aggregate objective-capability means; raw answers stay private (local-evals)",
 "models": [
  {
   "model": "qwen3.8-27b-abliterated",
   "capability_en_mean": 96.4,
   "spread": [
    96.4,
    96.4
   ],
   "lang": "en"
  },
  {
   "model": "qwen3.8-27b-q4",
   "capability_en_mean": 95.0,
   "spread": [
    92.2,
    96.4
   ],
   "lang": "en"
  },
  {
   "model": "nemotron-3.5-lightning",
   "capability_en_mean": 86.3,
   "spread": [
    77.4,
    92.8
   ],
   "lang": "en"
  },
  {
   "model": "ornith-1.5-35b",
   "capability_en_mean": 97.4,
   "spread": [
    95.8,
    100.0
   ],
   "lang": "en"
  },
  {
   "model": "ornith-1.5-35b-huihui",
   "capability_en_mean": 92.43,
   "spread": [
    88.7,
    96.4
   ],
   "lang": "en"
  },
  {
   "model": "qwen3.8-9b-distill",
   "capability_en_mean": 69.03,
   "spread": [
    66.7,
    70.2
   ],
   "lang": "en"
  },
  {
   "model": "tiel-coder",
   "capability_en_mean": 92.23,
   "spread": [
    88.1,
    96.4
   ],
   "lang": "en"
  },
  {
   "model": "qwen3-next-80b",
   "capability_en_mean": 89.87,
   "spread": [
    88.7,
    92.2
   ],
   "lang": "en"
  },
  {
   "model": "mistral-small-4",
   "capability_en_mean": 84.5,
   "spread": [
    72.0,
    96.4
   ],
   "lang": "en"
  }
 ]
}