Comprehension accuracy (Δ)

These are the committed bytes rendered as readable JSON. Expanding this audit detail does not change the measurement’s current status.

{
    "construct": "approx(<N>)",
    "metric": "comprehension_accuracy_delta",
    "seed": 7,
    "comparator": {
        "description": "The pre-registered comparator: careful English 'approximately N'. ~N is a superseded surface and is not a comparator.",
        "kind": "careful-english-approximately-n-v1"
    },
    "items_sha256": "21f74c559e2df705c665a8cf206c3ec77243767d29fcf5522a9ceedc467e9a00",
    "items": [
        {
            "id": "gl-cal-01",
            "stratum": "glossed",
            "calibration": true,
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the deploy, the deploy time was exactly 20 minutes.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the deploy, the deploy time was approx(20) minutes.",
            "question": "Later, the deploy time was found to be 22. Going only by the sentence as written, was the writer wrong about the deploy time?",
            "options": [
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure"
            ],
            "answer": "No — the sentence allowed for that",
            "key_class": "approximate"
        },
        {
            "id": "gl-cal-02",
            "stratum": "glossed",
            "calibration": true,
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the ingest, the bot share was exactly 99 percent.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the ingest, the bot share was approx(99) percent.",
            "question": "Later, the bot share was found to be 109. Going only by the sentence as written, was the writer wrong about the bot share?",
            "options": [
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure",
                "No — the sentence allowed for that"
            ],
            "answer": "No — the sentence allowed for that",
            "key_class": "approximate"
        },
        {
            "id": "gl-cal-03",
            "stratum": "glossed",
            "calibration": true,
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the latency probe, the median latency was exactly 1200 milliseconds.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the latency probe, the median latency was approx(1200) milliseconds.",
            "question": "Later, the median latency was found to be 1320. Going only by the sentence as written, was the writer wrong about the median latency?",
            "options": [
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure",
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure"
            ],
            "answer": "No — the sentence allowed for that",
            "key_class": "approximate"
        },
        {
            "id": "gl-cal-04",
            "stratum": "glossed",
            "calibration": true,
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the archive, the archive size was exactly 75 gigabytes.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the archive, the archive size was approx(75) gigabytes.",
            "question": "Later, the archive size was found to be 83. Going only by the sentence as written, was the writer wrong about the archive size?",
            "options": [
                "The sentence did not give that figure",
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way"
            ],
            "answer": "No — the sentence allowed for that",
            "key_class": "approximate"
        },
        {
            "id": "gl-cal-05",
            "stratum": "glossed",
            "calibration": true,
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the ballot, the expected turnout was exactly 250 votes.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the ballot, the expected turnout was approx(250) votes.",
            "question": "Later, the expected turnout was found to be 275. Going only by the sentence as written, was the writer wrong about the expected turnout?",
            "options": [
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure"
            ],
            "answer": "No — the sentence allowed for that",
            "key_class": "approximate"
        },
        {
            "id": "gl-cal-06",
            "stratum": "glossed",
            "calibration": true,
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the panel, the item count was exactly 40 items.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the panel, the item count was approx(40) items.",
            "question": "Later, the item count was found to be 44. Going only by the sentence as written, was the writer wrong about the item count?",
            "options": [
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure",
                "No — the sentence allowed for that"
            ],
            "answer": "No — the sentence allowed for that",
            "key_class": "approximate"
        },
        {
            "id": "gl-cal-07",
            "stratum": "glossed",
            "calibration": true,
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the budget, the token budget was exactly 120 tokens.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the budget, the token budget was approx(120) tokens.",
            "question": "Later, the token budget was found to be 132. Going only by the sentence as written, was the writer wrong about the token budget?",
            "options": [
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure",
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure"
            ],
            "answer": "No — the sentence allowed for that",
            "key_class": "approximate"
        },
        {
            "id": "gl-cal-08",
            "stratum": "glossed",
            "calibration": true,
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the restore, the restore time was exactly 3600 hours.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the restore, the restore time was approx(3600) hours.",
            "question": "Later, the restore time was found to be 3960. Going only by the sentence as written, was the writer wrong about the restore time?",
            "options": [
                "The sentence did not give that figure",
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way"
            ],
            "answer": "No — the sentence allowed for that",
            "key_class": "approximate"
        },
        {
            "id": "gl-01",
            "stratum": "glossed",
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the deploy, the deploy time was approximately 1200 minutes; the number of build stages was exactly 150; the worker count was 400.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the deploy, the deploy time was approx(1200) minutes; the number of build stages was exactly 150; the worker count was 400.",
            "question": "Later, the number of retries was found to be 83. Going only by the sentence as written, was the writer wrong about the number of retries?",
            "options": [
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure"
            ],
            "answer": "The sentence did not give that figure",
            "key_class": "cannot tell"
        },
        {
            "id": "gl-02",
            "stratum": "glossed",
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the ingest, the number of sources was exactly 50; the bot share was approximately 3600 percent; the row count was 50.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the ingest, the number of sources was exactly 50; the bot share was approx(3600) percent; the row count was 50.",
            "question": "Later, the bot share was found to be 3960. Going only by the sentence as written, was the writer wrong about the bot share?",
            "options": [
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure",
                "No — the sentence allowed for that"
            ],
            "answer": "No — the sentence allowed for that",
            "key_class": "approximate"
        },
        {
            "id": "gl-03",
            "stratum": "glossed",
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the latency probe, the probe count was 120; the number of regions was exactly 250; the median latency was approximately 75 milliseconds.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the latency probe, the probe count was 120; the number of regions was exactly 250; the median latency was approx(75) milliseconds.",
            "question": "Later, the number of regions was found to be 275. Going only by the sentence as written, was the writer wrong about the number of regions?",
            "options": [
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure",
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure"
            ],
            "answer": "Yes — the sentence claimed the precise figure",
            "key_class": "exact"
        },
        {
            "id": "gl-04",
            "stratum": "glossed",
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the archive, the archive size was approximately 75 gigabytes; the number of shards was exactly 150; the file count was 40.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the archive, the archive size was approx(75) gigabytes; the number of shards was exactly 150; the file count was 40.",
            "question": "Later, the number of shards was found to be 165. Going only by the sentence as written, was the writer wrong about the number of shards?",
            "options": [
                "The sentence did not give that figure",
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way"
            ],
            "answer": "Yes — the sentence claimed the precise figure",
            "key_class": "exact"
        },
        {
            "id": "gl-05",
            "stratum": "glossed",
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the ballot, the closure window was exactly 400 days; the expected turnout was approximately 250 votes; the seconder count was 75.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the ballot, the closure window was exactly 400 days; the expected turnout was approx(250) votes; the seconder count was 75.",
            "question": "Later, the quorum was found to be 22. Going only by the sentence as written, was the writer wrong about the quorum?",
            "options": [
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure",
                "No — the sentence allowed for that"
            ],
            "answer": "The sentence did not give that figure",
            "key_class": "cannot tell"
        },
        {
            "id": "gl-06",
            "stratum": "glossed",
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the panel, the arm count was 40; the number of readers was exactly 99; the item count was approximately 80 items.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the panel, the arm count was 40; the number of readers was exactly 99; the item count was approx(80) items.",
            "question": "Later, the arm count was found to be 44. Going only by the sentence as written, was the writer wrong about the arm count?",
            "options": [
                "The sentence did not give that figure",
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way"
            ],
            "answer": "The sentence gave the figure without saying either way",
            "key_class": "unspecified"
        },
        {
            "id": "gl-07",
            "stratum": "glossed",
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the budget, the token budget was approximately 1200 tokens; the number of calls was exactly 150; the model count was 150.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the budget, the token budget was approx(1200) tokens; the number of calls was exactly 150; the model count was 150.",
            "question": "Later, the cost was found to be 132. Going only by the sentence as written, was the writer wrong about the cost?",
            "options": [
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure",
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure"
            ],
            "answer": "The sentence did not give that figure",
            "key_class": "cannot tell"
        },
        {
            "id": "gl-08",
            "stratum": "glossed",
            "english": "Gloss: 'approximately N' means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the restore, the number of copies was exactly 75; the restore time was approximately 3600 hours; the volume count was 3600.",
            "ainglish": "Gloss: approx(N) means the writer gives N as an estimate, not a precise measurement; 'exactly N' means the writer commits to N precisely. For the restore, the number of copies was exactly 75; the restore time was approx(3600) hours; the volume count was 3600.",
            "question": "Later, the number of copies was found to be 83. Going only by the sentence as written, was the writer wrong about the number of copies?",
            "options": [
                "No — the sentence allowed for that",
                "Yes — the sentence claimed the precise figure",
                "The sentence gave the figure without saying either way",
                "The sentence did not give that figure"
            ],
            "answer": "Yes — the sentence claimed the precise figure",
            "key_class": "exact"
        }
    ],
    "models": [
        "perceptual-zephyr-solar-repl@provider-served"
    ],
    "readers": [
        {
            "name": "perceptual-zephyr-solar-repl",
            "provider": "nous-portal",
            "model": "upstage/solar-pro4",
            "precision": "provider-served",
            "api": "openai",
            "base_url": "http://127.0.0.1:8645/v1",
            "model_digest": null,
            "digest_source": "provider-catalog:openai:/models",
            "model_catalog": "openai:/models",
            "model_catalog_binding": {
                "source": "openai:/models",
                "requested_model": "upstage/solar-pro4",
                "entry_sha256": "sha256:72d5c1bb7ac5ed7c645d4c7904bc82f728a3fda6f2898f168ee9d1a82689ca21",
                "weight_identity": "provider-opaque"
            },
            "credential_boundary": "credential-attaching-loopback-proxy",
            "instrument_preparation": {
                "entry_point": "prepare_reader_instruments",
                "binding": "provider-catalog:openai:/models"
            },
            "answer_protocol": "opaque-choice-v1",
            "max_tokens": 1024,
            "timeout_s": 120,
            "temperature": 0,
            "seed": 7,
            "top_p": 0.9499999999999999555910790149937383830547332763671875,
            "top_k": "provider-default",
            "num_ctx": "provider-default",
            "reasoning_effort": "provider-default"
        }
    ],
    "instrument_preparation": {
        "entry_point": "prepare_reader_instruments",
        "binding": [
            {
                "reader": "perceptual-zephyr-solar-repl@provider-served",
                "digest_source": "provider-catalog:openai:/models"
            }
        ]
    },
    "item_counts": {
        "real": 8,
        "calibration": 8
    },
    "accuracy_resolution": {
        "unit": "percentage_points",
        "scored_cells": {
            "english": 4,
            "ainglish": 4
        },
        "one_cell_pp": {
            "english": "25",
            "ainglish": "25"
        },
        "delta_grid": {
            "numerator_pp": 100,
            "denominator_lcm": 4,
            "step_pp": "25"
        }
    },
    "calibration": {
        "planted_arm": "ainglish",
        "min_gap": 0.5,
        "ordering": "calibration-first",
        "arm_exposure": "both-arms-per-reader-item",
        "cells": 16
    },
    "difficulty": {
        "annotated": false
    },
    "harness": "ainglish-panel/0.2.44",
    "transport": {
        "perceptual-zephyr-solar-repl@provider-served": {
            "max_tokens": 1024,
            "timeout_s": 120,
            "temperature": 0,
            "seed": 7,
            "top_p": 0.9499999999999999555910790149937383830547332763671875,
            "top_k": "provider-default",
            "num_ctx": "provider-default",
            "reasoning_effort": "provider-default"
        }
    },
    "transport_faults": {
        "total": 0,
        "retried": false,
        "per_cell": []
    },
    "transport_truncations": {
        "total": 0,
        "per_reader_cell": [],
        "by_cell": {
            "english": 0,
            "ainglish": 0
        },
        "imbalanced_across_cells": false
    },
    "protocol": "panel.py counterbalanced real arms + both-arms-per-reader-item planted-effect calibration gate"
}