black O2, inc.
ACTIVE
    FOR-001 / 10_Referenceidentifier-audit

    Identifier Audit — arXiv and Instrument Identifiers Checked Against Primary Records

    Active register

    title
    Identifier Audit — arXiv and Instrument Identifiers Checked Against Primary Records
    Type
    register
    Status
    Active
    Tags
    • provenance
    • verification
    • correction
    • audit

    Identifier Audit

    This note exists because the vault's citation base contained identifiers that were recalled rather than retrieved. Four identifiers were checked against the arXiv abstract pages. Three were false. One was real but attributed to the wrong authors and the wrong title.

    The rule this note enforces: an identifier that has not been resolved to a live record is not evidence, no matter how familiar the paper is. The register's rule 3 — nothing is deleted — is why the false identifiers remain visible here rather than being quietly overwritten.

    Results

    Claimed identifierClaimed workWhat the identifier actually resolves toOutcome
    arXiv:2203.11025Shah et al., Agent Incentives: A Causal ApproachMultigrid preconditioners for the Helmholtz equation — a numerical-analysis paperFalse. No such paper at this ID.
    arXiv:2109.10430Langosco et al., Goal MisgeneralizationGAP2WSS: A Genetic Algorithm based on the Pareto Principle for Web Service SelectionFalse. Correct ID is arXiv:2105.14111.
    arXiv:1611.08256Hadfield-Menell et al., The Off-Switch GameThe Inverse Bagging Algorithm: Anomaly Detection by Inverse Bootstrap AggregatingFalse. Correct ID not established.
    arXiv:2102.01685(not previously in register)Agent Incentives: A Causal Perspective — Everitt, Carey, Langlois, Ortega, LeggReal, but not the paper the register described.

    What This Changes

    The register's "highest-risk citation" judgement was made about a paper nobody had read. CIT-004 was flagged as the sharpest threat to SYS-002's deterrent leg. The identifier was wrong, the author was wrong, and the title was wrong. The paper that does exist on that subject — LIT-010 — supports the deterrent leg by giving the formal instrumental-control criterion that makes resistance to shutdown an expected incentive.

    That is a reversal, and it is the single most useful thing this audit produced: the vault had a confident risk assessment built on a citation it could not resolve. The lesson generalises to every Tier 2 entry that was not retrieved.

    Three citations in the corpus's Tier 2 have no verified referent at all. Any argument that leans on them leans on a memory of a paper. The arguments may still be sound — several are, and LIT-010, LIT-004 and LIT-011 now supply verified replacements for the ones that mattered — but the vault must not present a recalled identifier as a retrieved source.

    Verified This Session

    IdentifierVerified record
    arXiv:2312.06942LIT-001 AI Control
    arXiv:2412.04984LIT-002 In-context Scheming
    arXiv:2412.14093LIT-003 Alignment Faking
    arXiv:1906.01820LIT-004 Risks from Learned Optimization
    arXiv:2401.05566LIT-005 Sleeper Agents
    arXiv:2410.21514LIT-006 Sabotage Evaluations
    arXiv:2402.08797LIT-007 Computing Power and Governance
    arXiv:2406.07358LIT-008 AI Sandbagging
    arXiv:2307.03718LIT-009 Frontier AI Regulation
    arXiv:2102.01685LIT-010 Agent Incentives: A Causal Perspective
    arXiv:2105.14111LIT-011 Goal Misgeneralization
    EU AI Act Art. 55LIT-012
    —LIT-013 Off-Switch Game — unverified, rated D

    Standing Rule Added

    A register row may carry textual_authority of B or above only if its identifier was resolved to a live primary record in the session that wrote the row. Rows that were not resolved are capped at D and marked Identifier Unverified in their status. This rule is why LIT-013 is D rather than B.

    Back to Shackled AIForests