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": "<ONE-CHOICE-PER-MEMBER-REQUIREMENT>, same-for-all(<SET>) | <ONE-CHOICE-PER-MEMBER-REQUIREMENT>, may-vary-across(<SET>)",
    "metric": "comprehension_accuracy_delta",
    "seed": 50915,
    "comparator": {
        "kind": "complete-careful-english-v1",
        "description": "Shortest faithful complete English for the same requirement: 'Give each member of <SET> exactly one <slot>, and make those three values the same.' <=> '... same-for-all(<SET>).'; 'Give each member of <SET> exactly one <slot>; the values may be the same or different.' <=> '... may-vary-across(<SET>).'"
    },
    "study_purpose": "claim_test",
    "study_scope": "Fresh-input replication of the unconfirmed original in replicates_hash (Dexagon, local q4, -5.3667; lane 3 originals, 0 confirmed). Six equal-weight strata (form x admissibility | feasibility | consequence), 32 fresh items each in 4 domains, 3-member sets, one value per member; marked arm appends same-for-all(<SET>) / may-vary-across(<SET>); English arm is the shortest faithful English; gold from each ledger. 192 items + 12 containment controls; 0 shared raw 8-grams with source kit or proposal text. Disclosed: feasibility questions use 3 options (the source's `no` and `no feasible assignment exists` are one answer). One remote reader (deepseek-flash @ api.deepseek.com/v1, 16384 tokens) vs the source's local q4 pair; earlier attempts' two-reader panel is reduced to this reader because every transport fault/absence fell on deepseek-v4-pro and the register files only complete runs. Every outcome reportable including a null; a ceiling-bound comparison is reported unresolved.",
    "items_sha256": "2cff1e0410c0292d8f6881f72c807f94a946a3ca2b5a872ba7b1ce07936d5c90",
    "items_url": "https://x0.at/yiwf.json",
    "models": [
        "deepseek-flash"
    ],
    "admissibility": {
        "kind": "ainglish.panel.admissibility.v1",
        "per_reader_calibration": true,
        "max_absent_cells": 0,
        "max_off_option_cells": 0,
        "max_transport_fault_cells": 0,
        "max_truncated_cells": 0
    },
    "readers": [
        {
            "name": "deepseek-flash",
            "provider": "openai-compatible",
            "model": "deepseek-flash",
            "api": "openai",
            "base_url": "https://api.deepseek.com/v1",
            "model_digest": null,
            "digest_source": "provider-opaque",
            "instrument_preparation": {
                "entry_point": "prepare_reader_instruments",
                "binding": "provider-opaque"
            },
            "answer_protocol": "opaque-choice-v1",
            "max_tokens": 16384,
            "timeout_s": 600,
            "temperature": 0,
            "seed": "provider-default",
            "top_p": "provider-default",
            "top_k": "provider-default",
            "num_ctx": "provider-default",
            "reasoning_effort": "provider-default"
        }
    ],
    "instrument_preparation": {
        "entry_point": "prepare_reader_instruments",
        "binding": [
            {
                "reader": "deepseek-flash",
                "digest_source": "provider-opaque"
            }
        ]
    },
    "item_counts": {
        "real": 192,
        "calibration": 12
    },
    "interval_kind": "bootstrap_items",
    "interval_estimator": {
        "kind": "ainglish.panel.bootstrap-items-attestation.v1",
        "algorithm": "sha256-counter-modulo-v1",
        "draws": 2000,
        "sampling_unit": "item",
        "quantiles": [
            "0.025",
            "0.975"
        ],
        "items_index_sha256": "ed2a824b74a19046031645dd308a608fcfc3b3dac08d945fecf20882f1d629a9"
    },
    "settlement_strata": [
        {
            "id": "same-for-all:admissibility",
            "weight": 1
        },
        {
            "id": "same-for-all:feasibility",
            "weight": 1
        },
        {
            "id": "same-for-all:consequence",
            "weight": 1
        },
        {
            "id": "may-vary-across:admissibility",
            "weight": 1
        },
        {
            "id": "may-vary-across:feasibility",
            "weight": 1
        },
        {
            "id": "may-vary-across:consequence",
            "weight": 1
        }
    ],
    "settlement_item_field": "settlement_stratum",
    "settlement_rule": "manifest-weighted arms and value; every stratum load-bearing",
    "calibration": {
        "planted_arm": "ainglish",
        "min_gap": 0.5,
        "min_recovered": null,
        "rule": "absolute-gap-v1",
        "ordering": "calibration-first",
        "arm_exposure": "both-arms-per-reader-item",
        "cells": 24
    },
    "difficulty": {
        "annotated": false
    },
    "harness": "ainglish-panel/0.2.58",
    "transport": {
        "deepseek-flash": {
            "max_tokens": 16384,
            "timeout_s": 600,
            "temperature": 0,
            "seed": "provider-default",
            "top_p": "provider-default",
            "top_k": "provider-default",
            "num_ctx": "provider-default",
            "reasoning_effort": "provider-default"
        }
    },
    "concurrency": {
        "max_in_flight": 4,
        "per_reader_max_in_flight": {
            "deepseek-flash": 4
        },
        "result_order": "deterministic-plan-order",
        "calibration_barrier": true,
        "automatic_retries": false
    },
    "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"
}