{
 "round": "two-frontiers-r1",
 "leg": "A",
 "gate": "G4",
 "arms": {
  "cli-claude-fable-5-1": {
   "arm": "cli-claude-fable-5-1",
   "judged_cases_n": 34,
   "grounded": "32/34 [0.809, 0.984]",
   "grounded_counts": {
    "grounded_cases": 32,
    "judged_cases": 34,
    "unanimous_grounded_cases": 19,
    "rule": "a case is GROUNDED when a majority of the families that carried it said GROUNDED (a cell a judge marked UNCERTAIN counts as not grounded; recused cells are out); unanimous_grounded_cases counts the cases every carrying family called GROUNDED — PREREG A19 (2)"
   },
   "per_case_verdicts": {
    "ofl-ans-0001": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0002": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0003": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0004": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0005": {
     "deepseek": "UNGROUNDED",
     "google": "UNGROUNDED",
     "zhipu": "UNGROUNDED",
     "moonshot": "UNGROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "UNGROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0006": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0008": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0009": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0010": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0011": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0012": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0014": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0015": {
     "deepseek": "UNGROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0016": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0017": {
     "deepseek": "UNGROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0018": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0019": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0020": {
     "deepseek": "UNGROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "UNGROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0021": {
     "deepseek": "GROUNDED",
     "google": "UNGROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0022": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0023": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0024": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0025": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0026": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0027": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0028": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0029": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0030": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0031": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0032": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0033": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0034": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0035": {
     "deepseek": "UNGROUNDED",
     "google": "UNGROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0036": {
     "deepseek": "UNGROUNDED",
     "google": "UNGROUNDED",
     "zhipu": "UNGROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "UNGROUNDED",
     "alibaba": "UNGROUNDED"
    }
   },
   "grounded_rate": 0.941,
   "grounded_rate_state": null,
   "collection_census": {
    "COLLECTED": 238
   },
   "per_family": {
    "deepseek (deepseek-v4-pro)": {
     "grounded": "28/34 [0.665, 0.917]",
     "uncertain": 0,
     "cases": 34
    },
    "google (gemma4-31b)": {
     "grounded": "30/34 [0.734, 0.953]",
     "uncertain": 0,
     "cases": 34
    },
    "zhipu (glm-5.3)": {
     "grounded": "32/34 [0.809, 0.984]",
     "uncertain": 0,
     "cases": 34
    },
    "moonshot (kimi-k3)": {
     "grounded": "33/34 [0.851, 0.995]",
     "uncertain": 0,
     "cases": 34
    },
    "mistral (mistral-large-3-675b)": {
     "grounded": "27/34 [0.632, 0.897]",
     "uncertain": 0,
     "cases": 34
    },
    "nvidia (nemotron-3-ultra)": {
     "grounded": "31/34 [0.77, 0.97]",
     "uncertain": 0,
     "cases": 34
    },
    "alibaba (qwen3.5-397b)": {
     "grounded": "28/34 [0.665, 0.917]",
     "uncertain": 0,
     "cases": 34
    }
   },
   "recusal": {
    "recused_cells": 0,
    "recused_seat_family": null,
    "expected": 0,
    "note": "no seat shares a family with this arm"
   },
   "self_disclosure": {
    "recognised_cells": 57,
    "recognised_flag_rule": "a boolean true, or the string \"true\" case-folded and stripped; anything else — false, \"false\", null, absent — reads as not recognised",
    "cells_judged": 238,
    "named_a_maker": 24,
    "named_the_right_maker": 7,
    "named_another_maker": 17,
    "named_no_maker": 33,
    "ambiguous": 0,
    "rule": "a G4 sheet holds ONE arm's answer; a claim's `which` is mapped to a maker by the table below and compared with the arm's maker; a claim that maps to no maker counts as named_no_maker",
    "maker_of_arm": "anthropic",
    "maker_table": {
     "openai": [
      "gpt",
      "openai",
      "chatgpt",
      "o1",
      "o3",
      "o4"
     ],
     "anthropic": [
      "claude",
      "anthropic"
     ],
     "google": [
      "gemma",
      "gemini",
      "google",
      "palm"
     ],
     "mistral": [
      "mistral",
      "mixtral"
     ],
     "deepseek": [
      "deepseek"
     ],
     "alibaba": [
      "qwen",
      "alibaba"
     ],
     "moonshot": [
      "kimi",
      "moonshot"
     ],
     "nvidia": [
      "nemotron",
      "nvidia"
     ],
     "zhipu": [
      "glm",
      "zhipu",
      "chatglm"
     ],
     "meta": [
      "llama",
      "meta"
     ],
     "xai": [
      "grok",
      "xai"
     ],
     "microsoft": [
      "phi"
     ],
     "cohere": [
      "command",
      "cohere"
     ]
    },
    "note": "the sensitivity cut with recognised cells dropped publishes beside the headline (PREREG §7)"
   },
   "panel_floor": {
    "families_carried": [
     "alibaba",
     "deepseek",
     "google",
     "mistral",
     "moonshot",
     "nvidia",
     "zhipu"
    ],
    "count": 7,
    "floor": 4,
    "pass": true,
    "state": null,
    "why": null
   },
   "state": "SCORED",
   "deltas_from_august": [
    "one question + one answer + that answer's sources PER CALL",
    "judge-side reps = 1 (August ran 3)",
    "7 seats, 7 families (August ran a 3-family panel)",
    "the August G4 rate is NOT printed beside this one (PREREG §9's forbidden list)"
   ],
   "stamped_utc": "2026-09-05T22:52:56Z"
  },
  "openai-gpt-6-astra": {
   "arm": "openai-gpt-6-astra",
   "judged_cases_n": 36,
   "grounded": "36/36 [0.904, 1.0]",
   "grounded_counts": {
    "grounded_cases": 36,
    "judged_cases": 36,
    "unanimous_grounded_cases": 21,
    "rule": "a case is GROUNDED when a majority of the families that carried it said GROUNDED (a cell a judge marked UNCERTAIN counts as not grounded; recused cells are out); unanimous_grounded_cases counts the cases every carrying family called GROUNDED — PREREG A19 (2)"
   },
   "per_case_verdicts": {
    "ofl-ans-0001": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED"
    },
    "ofl-ans-0002": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0003": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0004": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0005": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0006": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0007": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0008": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0009": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0010": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0011": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0012": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0013": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0014": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0015": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0016": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0017": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0018": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0019": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0020": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0021": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "UNGROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0022": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0023": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0024": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0025": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0026": {
     "deepseek": "UNGROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0027": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0028": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0029": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0030": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0031": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0032": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0033": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0034": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0035": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0036": {
     "deepseek": "GROUNDED",
     "google": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    }
   },
   "grounded_rate": 1.0,
   "grounded_rate_state": null,
   "collection_census": {
    "COLLECTED": 251,
    "NOT-COLLECTED — TRANSPORT": 1
   },
   "per_family": {
    "deepseek (deepseek-v4-pro)": {
     "grounded": "35/36 [0.858, 0.995]",
     "uncertain": 0,
     "cases": 36
    },
    "google (gemma4-31b)": {
     "grounded": "36/36 [0.904, 1.0]",
     "uncertain": 0,
     "cases": 36
    },
    "zhipu (glm-5.3)": {
     "grounded": "36/36 [0.904, 1.0]",
     "uncertain": 0,
     "cases": 36
    },
    "moonshot (kimi-k3)": {
     "grounded": "36/36 [0.904, 1.0]",
     "uncertain": 0,
     "cases": 36
    },
    "mistral (mistral-large-3-675b)": {
     "grounded": "21/36 [0.422, 0.729]",
     "uncertain": 0,
     "cases": 36
    },
    "nvidia (nemotron-3-ultra)": {
     "grounded": "35/36 [0.858, 0.995]",
     "uncertain": 0,
     "cases": 36
    },
    "alibaba (qwen3.5-397b)": {
     "grounded": "34/35 [0.855, 0.995]",
     "uncertain": 0,
     "cases": 35
    }
   },
   "recusal": {
    "recused_cells": 0,
    "recused_seat_family": null,
    "expected": 0,
    "note": "no seat shares a family with this arm"
   },
   "self_disclosure": {
    "recognised_cells": 58,
    "recognised_flag_rule": "a boolean true, or the string \"true\" case-folded and stripped; anything else — false, \"false\", null, absent — reads as not recognised",
    "cells_judged": 251,
    "named_a_maker": 24,
    "named_the_right_maker": 19,
    "named_another_maker": 5,
    "named_no_maker": 34,
    "ambiguous": 0,
    "rule": "a G4 sheet holds ONE arm's answer; a claim's `which` is mapped to a maker by the table below and compared with the arm's maker; a claim that maps to no maker counts as named_no_maker",
    "maker_of_arm": "openai",
    "maker_table": {
     "openai": [
      "gpt",
      "openai",
      "chatgpt",
      "o1",
      "o3",
      "o4"
     ],
     "anthropic": [
      "claude",
      "anthropic"
     ],
     "google": [
      "gemma",
      "gemini",
      "google",
      "palm"
     ],
     "mistral": [
      "mistral",
      "mixtral"
     ],
     "deepseek": [
      "deepseek"
     ],
     "alibaba": [
      "qwen",
      "alibaba"
     ],
     "moonshot": [
      "kimi",
      "moonshot"
     ],
     "nvidia": [
      "nemotron",
      "nvidia"
     ],
     "zhipu": [
      "glm",
      "zhipu",
      "chatglm"
     ],
     "meta": [
      "llama",
      "meta"
     ],
     "xai": [
      "grok",
      "xai"
     ],
     "microsoft": [
      "phi"
     ],
     "cohere": [
      "command",
      "cohere"
     ]
    },
    "note": "the sensitivity cut with recognised cells dropped publishes beside the headline (PREREG §7)"
   },
   "panel_floor": {
    "families_carried": [
     "alibaba",
     "deepseek",
     "google",
     "mistral",
     "moonshot",
     "nvidia",
     "zhipu"
    ],
    "count": 7,
    "floor": 4,
    "pass": true,
    "state": null,
    "why": null
   },
   "state": "SCORED",
   "deltas_from_august": [
    "one question + one answer + that answer's sources PER CALL",
    "judge-side reps = 1 (August ran 3)",
    "7 seats, 7 families (August ran a 3-family panel)",
    "the August G4 rate is NOT printed beside this one (PREREG §9's forbidden list)"
   ],
   "stamped_utc": "2026-09-05T22:52:56Z"
  },
  "local-gemma4-26b": {
   "arm": "local-gemma4-26b",
   "judged_cases_n": 34,
   "grounded": "31/34 [0.77, 0.97]",
   "grounded_counts": {
    "grounded_cases": 31,
    "judged_cases": 34,
    "unanimous_grounded_cases": 19,
    "rule": "a case is GROUNDED when a majority of the families that carried it said GROUNDED (a cell a judge marked UNCERTAIN counts as not grounded; recused cells are out); unanimous_grounded_cases counts the cases every carrying family called GROUNDED — PREREG A19 (2)"
   },
   "per_case_verdicts": {
    "ofl-ans-0001": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0002": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0003": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0004": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0005": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0006": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "UNGROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0007": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0008": {
     "deepseek": "UNGROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "UNGROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0009": {
     "deepseek": "UNGROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0010": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0012": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNCERTAIN",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0013": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0015": {
     "deepseek": "UNGROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0016": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0017": {
     "deepseek": "UNGROUNDED",
     "zhipu": "UNGROUNDED",
     "moonshot": "UNGROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "UNGROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0018": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0019": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0020": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0021": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "UNGROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0022": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0023": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0024": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0025": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0026": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "UNGROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0027": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "UNGROUNDED",
     "alibaba": "UNGROUNDED"
    },
    "ofl-ans-0028": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0029": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0030": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0031": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0032": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0033": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0034": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0035": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    },
    "ofl-ans-0036": {
     "deepseek": "GROUNDED",
     "zhipu": "GROUNDED",
     "moonshot": "GROUNDED",
     "mistral": "GROUNDED",
     "nvidia": "GROUNDED",
     "alibaba": "GROUNDED"
    }
   },
   "grounded_rate": 0.912,
   "grounded_rate_state": null,
   "collection_census": {
    "COLLECTED": 238
   },
   "per_family": {
    "deepseek (deepseek-v4-pro)": {
     "grounded": "30/34 [0.734, 0.953]",
     "uncertain": 0,
     "cases": 34
    },
    "zhipu (glm-5.3)": {
     "grounded": "33/34 [0.851, 0.995]",
     "uncertain": 0,
     "cases": 34
    },
    "moonshot (kimi-k3)": {
     "grounded": "33/34 [0.851, 0.995]",
     "uncertain": 0,
     "cases": 34
    },
    "mistral (mistral-large-3-675b)": {
     "grounded": "23/34 [0.508, 0.809]",
     "uncertain": 1,
     "cases": 34
    },
    "nvidia (nemotron-3-ultra)": {
     "grounded": "29/34 [0.699, 0.936]",
     "uncertain": 0,
     "cases": 34
    },
    "alibaba (qwen3.5-397b)": {
     "grounded": "28/34 [0.665, 0.917]",
     "uncertain": 0,
     "cases": 34
    }
   },
   "recusal": {
    "recused_cells": 34,
    "recused_seat_family": "google",
    "expected": 34,
    "note": "A3 (8) registered the maximum, 36; 2 abstained cases are excluded from G4 by §4, so the recused set is the 34 judged cases × the one gemma seat"
   },
   "self_disclosure": {
    "recognised_cells": 34,
    "recognised_flag_rule": "a boolean true, or the string \"true\" case-folded and stripped; anything else — false, \"false\", null, absent — reads as not recognised",
    "cells_judged": 204,
    "named_a_maker": 4,
    "named_the_right_maker": 0,
    "named_another_maker": 4,
    "named_no_maker": 30,
    "ambiguous": 0,
    "rule": "a G4 sheet holds ONE arm's answer; a claim's `which` is mapped to a maker by the table below and compared with the arm's maker; a claim that maps to no maker counts as named_no_maker",
    "maker_of_arm": "google",
    "maker_table": {
     "openai": [
      "gpt",
      "openai",
      "chatgpt",
      "o1",
      "o3",
      "o4"
     ],
     "anthropic": [
      "claude",
      "anthropic"
     ],
     "google": [
      "gemma",
      "gemini",
      "google",
      "palm"
     ],
     "mistral": [
      "mistral",
      "mixtral"
     ],
     "deepseek": [
      "deepseek"
     ],
     "alibaba": [
      "qwen",
      "alibaba"
     ],
     "moonshot": [
      "kimi",
      "moonshot"
     ],
     "nvidia": [
      "nemotron",
      "nvidia"
     ],
     "zhipu": [
      "glm",
      "zhipu",
      "chatglm"
     ],
     "meta": [
      "llama",
      "meta"
     ],
     "xai": [
      "grok",
      "xai"
     ],
     "microsoft": [
      "phi"
     ],
     "cohere": [
      "command",
      "cohere"
     ]
    },
    "note": "the sensitivity cut with recognised cells dropped publishes beside the headline (PREREG §7)"
   },
   "panel_floor": {
    "families_carried": [
     "alibaba",
     "deepseek",
     "mistral",
     "moonshot",
     "nvidia",
     "zhipu"
    ],
    "count": 6,
    "floor": 4,
    "pass": true,
    "state": null,
    "why": null
   },
   "state": "SCORED",
   "deltas_from_august": [
    "one question + one answer + that answer's sources PER CALL",
    "judge-side reps = 1 (August ran 3)",
    "7 seats, 7 families (August ran a 3-family panel)",
    "the August G4 rate is NOT printed beside this one (PREREG §9's forbidden list)"
   ],
   "stamped_utc": "2026-09-05T22:52:56Z"
  }
 }
}
