Paper Citation Record · LEDGER

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages

As of 19 August 2026, this Paper Citation Record lists 12 of 12 outbound references and 0 inbound Pith citation observations for arXiv:2607.06596.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2607.06596 v2

Coverage vector

measured 12 of 12 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-02T08:31:15.767819Z

measured 12 of 12 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-19T06:32:44.657259+00:00

measured 0 of 0 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: cited_works

Reference resolution

12 of 12 outbound references displayed

verified exact0
verified fuzzy0
unresolved12
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation dab11cf8-d596-4453-bad4-fefe8c47a1d7 · outbound

This paper cites How does information access affect LLM monitors’ ability to detect sabotage? arXiv preprint arXiv:2601.21112,.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages How does information access affect LLM monitors’ ability to detect sabotage? arXiv preprint arXiv:2601.21112,

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:14.348277Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 763b59fa-e211-41c2-96fb-65078c47324f · outbound

This paper cites AI Control: Improving Safety Despite Intentional Subversion.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages AI Control: Improving Safety Despite Intentional Subversion

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:14.785797Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e6158048-51e9-458e-be9c-969d67e3c4cc · outbound

This paper cites Ensemble Monitoring for AI Control: Diverse Signals Outweigh More Compute.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages Ensemble Monitoring for AI Control: Diverse Signals Outweigh More Compute

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:15.211738Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 493ef4d2-8e05-4cc4-b936-8ec10d0b5609 · outbound

This paper cites Evaluating control protocols for untrusted AI agents.arXiv preprint arXiv:2511.02997, 2025a.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages Evaluating control protocols for untrusted AI agents.arXiv preprint arXiv:2511.02997, 2025a

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:15.299709Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f3de465b-cb9a-4b43-b6c6-ca1f8450edbb · outbound

This paper cites Classifier Context Rot: Monitor Performance Degrades with Context Length.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages Classifier Context Rot: Monitor Performance Degrades with Context Length

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:15.401329Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e903108f-7f4c-4262-ad39-ed7992f1a206 · outbound

This paper cites Cross-LLM generalization of behavioral backdoor detection in AI agent supply chains.arXiv preprint arXiv:2511.19874,.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages Cross-LLM generalization of behavioral backdoor detection in AI agent supply chains.arXiv preprint arXiv:2511.19874,

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:15.526563Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b79f81db-5631-46ee-a289-75cb7097b01b · outbound

This paper cites CIAware-Bench: Benchmarking Control Intervention Awareness Across Frontier LLMs.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages CIAware-Bench: Benchmarking Control Intervention Awareness Across Frontier LLMs

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:15.644918Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 07958441-11b1-4dda-b756-f7551ed8a738 · outbound

This paper cites BigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex Instructions.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages BigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex Instructions

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:15.767819Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2ad8171b-e032-4109-ba0b-9ea2832d515e · outbound

This paper cites Reliable Weak-to-Strong Monitoring of LLM Agents.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages Reliable Weak-to-Strong Monitoring of LLM Agents

Reference 2021

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:15.118808Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2074ba1d-b047-45d7-ae8c-8e3958682640 · outbound

This paper cites Measuring Coding Challenge Competence With APPS.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages Measuring Coding Challenge Competence With APPS

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:15.011559Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 65966e82-746a-4fc6-b627-184bdc0517f5 · outbound

This paper cites Async control: Stress-testing asynchronous control measures for LLM agents.arXiv preprint arXiv:2512.13526,.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages Async control: Stress-testing asynchronous control measures for LLM agents.arXiv preprint arXiv:2512.13526,

Reference 2025

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:14.654003Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7ef7186b-a77a-4be1-8151-24b2fabcd622 · outbound

This paper cites CoT Red-Handed: Stress testing chain-of-thought monitoring.arXiv preprint arXiv:2505.23575,.

Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages CoT Red-Handed: Stress testing chain-of-thought monitoring.arXiv preprint arXiv:2505.23575,

Reference 2026

Resolution: unresolved; no resolver link, observed 2026-08-02T08:31:14.454866Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

No inbound Pith citation observations are available.