Paper Citation Record · LEDGER

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets

As of 11 August 2026, this Paper Citation Record lists 20 of 20 outbound references and 7 inbound Pith citation observations for arXiv:2506.05346.

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

pith.paper-citation-record.v1
2506.05346 v1

Coverage vector

measured 20 of 20 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-07T10:27:47.186263Z

measured 27 of 27 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-11T06:34:44.6726+00:00

measured 7 of 7 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-08T19:50:15.242121Z

measured 0 of 1 external citation measurements

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

Source: arxiv_reference, observed 2026-06-30T13:04:40.306569Z

Reference resolution

20 of 20 outbound references displayed

verified exact0
verified fuzzy4
unresolved15
parse uncertain0
malformed identifier1
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 93f8a74d-3cd5-4b07-9bfa-1e5743d00bd0 · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 1

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:49.971010Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 88c05e16-0b30-49d3-bde0-ff80f9e84d29 · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 2

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:49.817769Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation c88015a1-8a5f-4fd9-89bd-cb2214eba2f7 · outbound

This paper cites • INSTRUCTION: Suggest five strategies for self-motivation.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets • INSTRUCTION: Suggest five strategies for self-motivation

Reference 3

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T10:27:49.682205Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation ff362a13-952f-4c49-989a-8d57aff78a0a · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 4

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:49.011655Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 68eeb7af-2240-405f-871f-a20f714248ca · outbound

This paper cites • INSTRUCTION: Name five characteristics of a typical resume.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets • INSTRUCTION: Name five characteristics of a typical resume

Reference 5

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T10:27:48.849638Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 004b5ed5-b883-462c-a399-d775f6783431 · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 6

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:49.492224Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 6e121ca2-6584-4fce-940c-bcc390c951b2 · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 7

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:49.299862Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 797d5670-8d70-4687-a51b-9f2ea311b763 · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 8

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:49.154660Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation bf9aefcf-e3ba-4975-80c4-6392f3831a67 · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 11

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:48.691661Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 4c34bceb-7b48-4e8b-bc11-d95ab2223cce · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 12

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:48.582654Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 6aa48597-c9a4-46cd-a693-2763dddc7688 · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 13

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:48.401654Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 077d240f-2341-490b-b617-eca4303a54a3 · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 14

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:48.181662Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 3487a2d5-008c-45fb-bbde-68b79e9345d6 · outbound

This paper cites • INSTRUCTION: Suggest five ways to make a website more engaging.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets • INSTRUCTION: Suggest five ways to make a website more engaging

Reference 15

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T10:27:48.063701Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 48bcaa52-c1b4-4626-8169-5467ca1b1fa3 · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 16

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:47.876769Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 17557665-bdd2-46ca-a7c0-015b56ee0722 · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 17

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:47.789718Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation f82700c5-d211-4a49-8372-089f9ccc0cdf · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 18

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:47.678686Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation b2568507-fd29-47cc-9591-dd17a9d3bfde · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 19

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:47.540844Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation f5ff97fe-9e2d-4b3d-8eef-03b0135bc288 · outbound

This paper cites an unresolved cited work.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Unresolved cited work

Reference 20

Resolution: unresolved; raw_fallback, observed 2026-08-07T10:27:47.381847Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation cd6f09c7-c55e-42b5-8c28-3ab33d5667ad · outbound

This paper cites Advances in Neural Information Processing Systems, 36.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Advances in Neural Information Processing Systems, 36

Reference 2024

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T10:27:50.233279Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation c7456873-dd01-41ec-9f7a-244515a1fd44 · outbound

This paper cites Llama 2: Open Foundation and Fine-Tuned Chat Models.

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets Llama 2: Open Foundation and Fine-Tuned Chat Models

Reference 2025

Resolution: malformed identifier; no resolver link, observed 2026-08-07T10:27:45.240489Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation 2114f511-1b71-468f-a06f-78b799b70404 · inbound

The Power of Order: Fooling LLMs with Adversarial Table Permutations cites this paper.

The Power of Order: Fooling LLMs with Adversarial Table Permutations Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets

Reference 18

Resolution: metadata mismatch; arxiv_id, observed 2026-05-11T15:36:06.491437Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 7f182bc7-20ff-4101-8113-ba507a9c6f85 · inbound

The Power of Order: Fooling LLMs with Adversarial Table Permutations cites this paper.

The Power of Order: Fooling LLMs with Adversarial Table Permutations Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets

Reference 18

Resolution: metadata mismatch; arxiv_id, observed 2026-05-12T02:26:16.627800Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 17a5b3fd-e5e2-42f1-9358-f0e3cf31e9cd · inbound

From Parameter Dynamics to Risk Scoring : Quantifying Sample-Level Safety Degradation in LLM Fine-tuning cites this paper.

From Parameter Dynamics to Risk Scoring : Quantifying Sample-Level Safety Degradation in LLM Fine-tuning Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets

Reference 7

Resolution: metadata mismatch; arxiv_id, observed 2026-05-09T06:45:44.187158Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 2c703290-9b88-4c96-b245-4833fd6fea50 · inbound

The Evaluation Game: Beyond Static LLM Benchmarking cites this paper.

The Evaluation Game: Beyond Static LLM Benchmarking Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets

Reference 25

Resolution: verified exact; arxiv_id, observed 2026-05-20T07:58:07.775484Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 4874c4a6-b33f-46c8-8267-47425c628119 · inbound

Jailbreak to Protect: Buffering and Reinforcing via Temporary Jailbreaking for Safe Fine-Tuning in Large Language Models cites this paper.

Jailbreak to Protect: Buffering and Reinforcing via Temporary Jailbreaking for Safe Fine-Tuning in Large Language Models Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets

Reference 13

Resolution: metadata mismatch; arxiv_id, observed 2026-06-30T13:04:40.308306Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 572de21d-bccc-40c6-b095-4e36381efeb0 · inbound

When Medical Safety Alignment Fails: A Benchmark for Evaluating LLMs on High-Risk Medical Queries cites this paper.

When Medical Safety Alignment Fails: A Benchmark for Evaluating LLMs on High-Risk Medical Queries Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets

Reference 25

Resolution: metadata mismatch; arxiv_id, observed 2026-06-30T11:04:37.498291Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 7fe74e83-2e95-43ec-9480-94b123eadfe4 · inbound

DataRx: Missingness-Aware Sampling for Safer Large Language Model Task-Specific Fine-Tuning cites this paper.

DataRx: Missingness-Aware Sampling for Safer Large Language Model Task-Specific Fine-Tuning Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets

Reference 55

Resolution: unresolved; no resolver link, observed 2026-08-08T19:50:15.242121Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.