{
  "methods": [
    {
      "method_id": "autoscientists",
      "cells": {
        "params": {
          "rank": 2,
          "value": 13910060,
          "run_id": "autoscientists--params",
          "report_url": null
        },
        "flops": {
          "rank": 1,
          "value": 52148760,
          "run_id": "autoscientists--flops",
          "report_url": null
        },
        "memory": {
          "rank": 3,
          "value": 559877120,
          "run_id": "autoscientists--memory",
          "report_url": null
        },
        "traintime": {
          "rank": 2,
          "value": 240,
          "run_id": "autoscientists--traintime",
          "report_url": null
        },
        "steps": {
          "rank": 1,
          "value": 267,
          "run_id": "autoscientists--steps",
          "report_url": null
        },
        "data": {
          "rank": 5,
          "value": 184549376,
          "run_id": "autoscientists--data",
          "report_url": null
        },
        "kvcache": {
          "rank": 2,
          "value": 59392,
          "run_id": "autoscientists--kvcache",
          "report_url": null
        },
        "decode": {
          "rank": 1,
          "value": 49.695491790771484,
          "run_id": "autoscientists--decode",
          "report_url": null
        },
        "request": {
          "rank": 1,
          "value": 29.457807540893555,
          "run_id": "autoscientists--request",
          "report_url": null
        }
      },
      "coverage": 9,
      "matches": 54,
      "elo": 1729.3701786268632,
      "match_order_sd": 16.007790631060917,
      "rank": 1
    },
    {
      "method_id": "sequential-search",
      "cells": {
        "params": {
          "rank": 5,
          "value": 15151312,
          "run_id": "sequential-search--params",
          "report_url": null
        },
        "flops": {
          "rank": 2,
          "value": 56663808,
          "run_id": "sequential-search--flops",
          "report_url": null
        },
        "memory": {
          "rank": 4,
          "value": 608924672,
          "run_id": "sequential-search--memory",
          "report_url": null
        },
        "traintime": {
          "rank": 1,
          "value": 210,
          "run_id": "sequential-search--traintime",
          "report_url": null
        },
        "steps": {
          "rank": 2,
          "value": 295,
          "run_id": "sequential-search--steps",
          "report_url": null
        },
        "data": {
          "rank": 1,
          "value": 152455168,
          "run_id": "sequential-search--data",
          "report_url": null
        },
        "kvcache": {
          "rank": 1,
          "value": 1024,
          "run_id": "sequential-search--kvcache",
          "report_url": null
        },
        "decode": {
          "rank": 6,
          "value": 98.5555648803711,
          "run_id": "sequential-search--decode",
          "report_url": null
        },
        "request": {
          "rank": 3,
          "value": 61.29932403564453,
          "run_id": "sequential-search--request",
          "report_url": null
        }
      },
      "coverage": 9,
      "matches": 54,
      "elo": 1687.6113691095902,
      "match_order_sd": 17.939556954840572,
      "rank": 2
    },
    {
      "method_id": "arbor",
      "cells": {
        "params": {
          "rank": 4,
          "value": 14549136,
          "run_id": "arbor--params",
          "report_url": null
        },
        "flops": {
          "rank": 3,
          "value": 58296576,
          "run_id": "arbor--flops",
          "report_url": null
        },
        "memory": {
          "rank": 1,
          "value": 400084992,
          "run_id": "arbor--memory",
          "report_url": null
        },
        "traintime": {
          "rank": 3,
          "value": 270,
          "run_id": "arbor--traintime",
          "report_url": null
        },
        "steps": {
          "rank": 5,
          "value": 337,
          "run_id": "arbor--steps",
          "report_url": null
        },
        "data": {
          "rank": 3,
          "value": 159776768,
          "run_id": "arbor--data",
          "report_url": null
        },
        "kvcache": {
          "rank": 4,
          "value": 96768,
          "run_id": "arbor--kvcache",
          "report_url": null
        },
        "decode": {
          "rank": 3,
          "value": 57.08944797515869,
          "run_id": "arbor--decode",
          "report_url": null
        },
        "request": {
          "rank": 2,
          "value": 57.88278579711914,
          "run_id": "arbor--request",
          "report_url": null
        }
      },
      "coverage": 9,
      "matches": 54,
      "elo": 1632.374102089888,
      "match_order_sd": 12.447135436229603,
      "rank": 3
    },
    {
      "method_id": "gear",
      "cells": {
        "params": {
          "rank": 1,
          "value": 13197712,
          "run_id": "gear--params",
          "report_url": null
        },
        "flops": {
          "rank": 4,
          "value": 62233344,
          "run_id": "gear--flops",
          "report_url": null
        },
        "memory": {
          "rank": 2,
          "value": 466297856,
          "run_id": "gear--memory",
          "report_url": null
        },
        "traintime": {
          "rank": 3,
          "value": 270,
          "run_id": "gear--traintime",
          "report_url": null
        },
        "steps": {
          "rank": 3,
          "value": 306,
          "run_id": "gear--steps",
          "report_url": null
        },
        "data": {
          "rank": 4,
          "value": 164429824,
          "run_id": "gear--data",
          "report_url": null
        },
        "kvcache": {
          "rank": 5,
          "value": 124416,
          "run_id": "gear--kvcache",
          "report_url": null
        },
        "decode": {
          "rank": 2,
          "value": 56.014180183410645,
          "run_id": "gear--decode",
          "report_url": null
        },
        "request": {
          "rank": 4,
          "value": 97.84162044525146,
          "run_id": "gear--request",
          "report_url": null
        }
      },
      "coverage": 9,
      "matches": 54,
      "elo": 1622.4111446669015,
      "match_order_sd": 13.475478386820553,
      "rank": 4
    },
    {
      "method_id": "beam-search",
      "cells": {
        "params": {
          "rank": 3,
          "value": 14221456,
          "run_id": "beam-search--params",
          "report_url": null
        },
        "flops": {
          "rank": 5,
          "value": 69847872,
          "run_id": "beam-search--flops",
          "report_url": null
        },
        "memory": {
          "rank": 5,
          "value": 1020860928,
          "run_id": "beam-search--memory",
          "report_url": null
        },
        "traintime": {
          "rank": 3,
          "value": 270,
          "run_id": "beam-search--traintime",
          "report_url": null
        },
        "steps": {
          "rank": 4,
          "value": 335,
          "run_id": "beam-search--steps",
          "report_url": null
        },
        "data": {
          "rank": 6,
          "value": 197656576,
          "run_id": "beam-search--data",
          "report_url": null
        },
        "kvcache": {
          "rank": 3,
          "value": 69632,
          "run_id": "beam-search--kvcache",
          "report_url": null
        },
        "decode": {
          "rank": 4,
          "value": 65.14537334442139,
          "run_id": "beam-search--decode",
          "report_url": null
        },
        "request": {
          "rank": 6,
          "value": 180.55975437164307,
          "run_id": "beam-search--request",
          "report_url": null
        }
      },
      "coverage": 9,
      "matches": 54,
      "elo": 1438.6132957487964,
      "match_order_sd": 13.550908028890996,
      "rank": 5
    },
    {
      "method_id": "heuresis-map-elites",
      "cells": {
        "params": {
          "rank": 6,
          "value": 17826840,
          "run_id": "heuresis-map-elites--params",
          "report_url": null
        },
        "flops": {
          "rank": 6,
          "value": 84154368,
          "run_id": "heuresis-map-elites--flops",
          "report_url": null
        },
        "memory": {
          "rank": 6,
          "value": 1698522112,
          "run_id": "heuresis-map-elites--memory",
          "report_url": null
        },
        "traintime": {
          "rank": 3,
          "value": 270,
          "run_id": "heuresis-map-elites--traintime",
          "report_url": null
        },
        "steps": {
          "rank": 6,
          "value": 377,
          "run_id": "heuresis-map-elites--steps",
          "report_url": null
        },
        "data": {
          "rank": 2,
          "value": 156172288,
          "run_id": "heuresis-map-elites--data",
          "report_url": null
        },
        "kvcache": {
          "rank": 6,
          "value": 500224,
          "run_id": "heuresis-map-elites--kvcache",
          "report_url": null
        },
        "decode": {
          "rank": 5,
          "value": 97.81789779663086,
          "run_id": "heuresis-map-elites--decode",
          "report_url": null
        },
        "request": {
          "rank": 5,
          "value": 129.0278434753418,
          "run_id": "heuresis-map-elites--request",
          "report_url": null
        }
      },
      "coverage": 9,
      "matches": 54,
      "elo": 1350.4494901637602,
      "match_order_sd": 14.18883770983771,
      "rank": 6
    },
    {
      "method_id": "tpe",
      "cells": {
        "params": {
          "rank": 7,
          "value": 26345772,
          "run_id": "tpe--params",
          "report_url": null
        },
        "flops": {
          "rank": 7,
          "value": 100271808,
          "run_id": "tpe--flops",
          "report_url": null
        },
        "memory": {
          "rank": 7,
          "value": 6387505664,
          "run_id": "tpe--memory",
          "report_url": null
        },
        "traintime": {
          "rank": 7,
          "value": 510,
          "run_id": "tpe--traintime",
          "report_url": null
        },
        "steps": {
          "rank": 7,
          "value": 753,
          "run_id": "tpe--steps",
          "report_url": null
        },
        "data": {
          "rank": 7,
          "value": 387710976,
          "run_id": "tpe--data",
          "report_url": null
        },
        "kvcache": {
          "rank": 7,
          "value": 18219520,
          "run_id": "tpe--kvcache",
          "report_url": null
        },
        "decode": {
          "rank": 7,
          "value": 230.74018955230713,
          "run_id": "tpe--decode",
          "report_url": null
        },
        "request": {
          "rank": 7,
          "value": 295.38750648498535,
          "run_id": "tpe--request",
          "report_url": null
        }
      },
      "coverage": 9,
      "matches": 54,
      "elo": 1039.1704195942107,
      "match_order_sd": 6.855684393795971,
      "rank": 7
    }
  ],
  "status": "ready",
  "task_count": 9,
  "comparison_count": 189,
  "rating": {
    "method": "Order-averaged margin-aware Elo",
    "base": 1500,
    "scale": 400,
    "cohort_mean": 1500,
    "leader_anchor": null,
    "shuffles": 4000,
    "seed": 20260902,
    "k_factor": 32.0,
    "method_order": [
      "sequential-search",
      "beam-search",
      "autoscientists",
      "arbor",
      "gear",
      "heuresis-map-elites",
      "tpe"
    ],
    "target_order": [
      "params",
      "flops",
      "memory",
      "traintime",
      "steps",
      "data",
      "kvcache",
      "decode",
      "request"
    ],
    "weighting": "equal_per_included_target",
    "target_weights": {
      "params": 0.1111111111111111,
      "flops": 0.1111111111111111,
      "memory": 0.1111111111111111,
      "traintime": 0.1111111111111111,
      "steps": 0.1111111111111111,
      "data": 0.1111111111111111,
      "kvcache": 0.1111111111111111,
      "decode": 0.1111111111111111,
      "request": 0.1111111111111111
    },
    "rounding_steps": {
      "traintime": 30
    },
    "rounding_formula": "step * floor(raw_value / step + 0.5)",
    "margin": "min(3, log1p(abs(log(right / left)) / target_scale) / log(2))",
    "target_scale": "median nonzero absolute pairwise log gap, floored at 1e-9; 1 if all tied",
    "comparison_order": "configured target and method order; new methods appended by ID",
    "run_selection": "best accepted run per method and task",
    "missing_targets": "excluded from comparisons",
    "ties": "equal target values: outcome 0.5, zero margin and zero update",
    "match_order_sd": "population standard deviation across shuffled passes; order sensitivity"
  }
}