Paper Citation Record · LEDGER

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents

As of 7 August 2026, this Paper Citation Record lists 15 of 15 outbound references and 0 inbound Pith citation observations for arXiv:2607.19449.

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

pith.paper-citation-record.v1
2607.19449 v1

Coverage vector

measured 15 of 15 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-01T13:32:29.381847Z

measured 15 of 15 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-07T06:34:17.273281+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

15 of 15 outbound references displayed

verified exact0
verified fuzzy0
unresolved15
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 18dd4466-8d49-4712-b031-27e77478d38b · outbound

This paper cites an unresolved cited work.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents Unresolved cited work

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:28.256476Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bd1b231e-7fde-42ba-9674-4e06797bdbb8 · outbound

This paper cites 2022.LangChain.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents 2022.LangChain

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:28.412560Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6ad12b20-9788-438a-8cf0-e574c99615c3 · outbound

This paper cites The Llama 3 Herd of Models.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents The Llama 3 Herd of Models

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:28.555493Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7dcb1b53-9b6e-46c7-8517-7893b3e65c5a · outbound

This paper cites Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:28.638390Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 97fb5f10-7320-4508-911d-19fc2702a96a · outbound

This paper cites an unresolved cited work.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents Unresolved cited work

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:28.698688Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4e784f80-2647-493d-8b47-f2a4b97f77fa · outbound

This paper cites an unresolved cited work.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents Unresolved cited work

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:28.754948Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1e0e3b43-9f07-43af-828d-bfca0611122a · outbound

This paper cites AgentBench: Evaluating LLMs as Agents.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents AgentBench: Evaluating LLMs as Agents

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:28.851035Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fcb696dc-df53-466f-bb80-3cafc00c1e03 · outbound

This paper cites GPT-4 Technical Report.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents GPT-4 Technical Report

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:28.941780Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6df75736-952d-45a1-a109-65ce7453fbf8 · outbound

This paper cites Toolformer: Language Models Can Teach Themselves to Use Tools.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents Toolformer: Language Models Can Teach Themselves to Use Tools

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:28.984369Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5d372112-1359-407f-a96a-bbd85947bd62 · outbound

This paper cites Towards Understanding Sycophancy in Language Models.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents Towards Understanding Sycophancy in Language Models

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:29.021850Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7198e472-2e76-4e26-8d90-da064b4f76cc · outbound

This paper cites Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:29.076683Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1afc3dad-63fd-47c2-b1e8-46643423f7e0 · outbound

This paper cites an unresolved cited work.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents Unresolved cited work

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:29.116487Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1672d1f7-f840-4c5a-88d9-0909195994ad · outbound

This paper cites SafeToolBench: Pioneering a Prospective Benchmark to Evaluating Tool Utilization Safety in LLMs.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents SafeToolBench: Pioneering a Prospective Benchmark to Evaluating Tool Utilization Safety in LLMs

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:29.176874Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 75d6808a-4b51-4bb6-a7fb-428db49446be · outbound

This paper cites ReAct: Synergizing Reasoning and Acting in Language Models.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents ReAct: Synergizing Reasoning and Acting in Language Models

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:29.283897Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 864a4ce6-ffbc-4913-b6a3-d864a99c8d07 · outbound

This paper cites the tool returned no data.

Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents the tool returned no data

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-01T13:32:29.381847Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

No inbound Pith citation observations are available.