Paper Citation Record · LEDGER

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak

As of 14 August 2026, this Paper Citation Record lists 44 of 44 outbound references and 0 inbound Pith citation observations for arXiv:2607.14147.

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

pith.paper-citation-record.v1
2607.14147 v1

Coverage vector

measured 44 of 44 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-02T06:30:57.585806Z

measured 44 of 44 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-14T06:32:32.682623+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

44 of 44 outbound references displayed

verified exact0
verified fuzzy0
unresolved44
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation df200c88-f912-46fe-9c1a-cf8ee85f9bcd · outbound

This paper cites International Conference on Learning Representations (ICLR) , year =.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak International Conference on Learning Representations (ICLR) , year =

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:51.484354Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9d0e9505-e038-4ef6-9c7f-77710fd6ff79 · outbound

This paper cites Advances in Neural Information Processing Systems (NeurIPS) , year =.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Advances in Neural Information Processing Systems (NeurIPS) , year =

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:51.565855Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 15fcd98a-b060-4801-8af1-cc49e43e8891 · outbound

This paper cites Advances in Neural Information Processing Systems (NeurIPS) , year =.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Advances in Neural Information Processing Systems (NeurIPS) , year =

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:51.857128Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9bb78f12-f9f3-4ce2-be76-08e0e9b3376e · outbound

This paper cites and Guo, K.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak and Guo, K

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:52.210431Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5647f612-1ef6-4037-8356-3e471cd1d3d4 · outbound

This paper cites BlackboxNLP Workshop , year =.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak BlackboxNLP Workshop , year =

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:52.586943Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d3db1c2e-74c4-4ce7-ae4a-38abe2b86bd3 · outbound

This paper cites Advances in Neural Information Processing Systems (NeurIPS) , year =.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Advances in Neural Information Processing Systems (NeurIPS) , year =

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:52.729064Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7d2c4e6f-3470-45de-a6cd-c74dcef5d52b · outbound

This paper cites NAACL , year =.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak NAACL , year =

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:52.791346Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b3dd62ce-13ad-4363-8759-cca3d4f8b9f4 · outbound

This paper cites 2026 , howpublished =.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak 2026 , howpublished =

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:52.859621Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c5675ead-84e6-4952-b418-9c4bdd0c6585 · outbound

This paper cites , year =.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak , year =

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:53.224053Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 63ce70c7-27a6-45a8-a09d-39ede7f6888a · outbound

This paper cites Catastrophic Jailbreak of Open-source.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Catastrophic Jailbreak of Open-source

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:53.302184Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d716021f-8fb1-4964-a39d-aac893415aed · outbound

This paper cites Think You Have Solved Question Answering? Try.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Think You Have Solved Question Answering? Try

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:53.422009Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b7e9a1de-4f94-4f60-b462-474168c1c2d2 · outbound

This paper cites Which Agent Causes Task Failures and When? On Automated Failure Attribution of.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Which Agent Causes Task Failures and When? On Automated Failure Attribution of

Reference 27

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:53.528719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 507b4e42-50fd-457e-ab8d-21744068cc00 · outbound

This paper cites Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations , pages =.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations , pages =

Reference 28

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:53.690667Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8b4013b2-691d-443a-8178-b36d08622fbb · outbound

This paper cites Advances in Neural Information Processing Systems (NeurIPS) , year =.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Advances in Neural Information Processing Systems (NeurIPS) , year =

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:53.857418Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 44b7fa37-8b3a-46c7-a919-c9bd54058df4 · outbound

This paper cites Journal of Machine Learning Research , volume =.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Journal of Machine Learning Research , volume =

Reference 30

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:53.944415Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8b50241c-85c4-4de9-b334-12ccbdb6ccc4 · outbound

This paper cites Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone

Reference 31

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:54.033545Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 923f07c8-eff0-4ca8-9b1d-82c03acb122f · outbound

This paper cites SmolLM2: When Smol Goes Big -- Data-Centric Training of a Small Language Model.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak SmolLM2: When Smol Goes Big -- Data-Centric Training of a Small Language Model

Reference 32

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:54.093344Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fb22f92b-2d6e-4398-ad1a-8fd1a1793758 · outbound

This paper cites Refusal in Language Models Is Mediated by a Single Direction.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Refusal in Language Models Is Mediated by a Single Direction

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:54.196799Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ca5aadf8-cf17-4b91-9b66-a7519657a5d8 · outbound

This paper cites What features in prompts jailbreak llms? investigating the mechanisms behind attacks.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak What features in prompts jailbreak llms? investigating the mechanisms behind attacks

Reference 34

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:54.296338Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 415a905c-23db-48d8-a56d-cbb56e48baff · outbound

This paper cites Before the Last Token: Diagnosing Final-Token Safety Probe Failures.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Before the Last Token: Diagnosing Final-Token Safety Probe Failures

Reference 35

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:54.427031Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 20f6e9c9-cbc4-4d71-a8c2-d1936b93d2b6 · outbound

This paper cites Arcanum prompt injection taxonomy.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Arcanum prompt injection taxonomy

Reference 36

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:54.594655Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1d592932-b1f2-4557-a19f-940be68fc7d7 · outbound

This paper cites an unresolved cited work.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Unresolved cited work

Reference 37

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:54.769346Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1b365633-993d-4c42-92e2-0b4d58f41721 · outbound

This paper cites Catastrophic Jailbreak of Open-source LLMs via Exploiting Generation.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Catastrophic Jailbreak of Open-source LLMs via Exploiting Generation

Reference 38

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:54.931280Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7f491374-b78d-4e27-a488-091acdda10af · outbound

This paper cites Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations

Reference 39

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:55.091334Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation be50c1f4-ba70-4939-be24-6620ae2bb353 · outbound

This paper cites HiddenDetect: Detecting Jailbreak Attacks against Large Vision-Language Models via Monitoring Hidden States.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak HiddenDetect: Detecting Jailbreak Attacks against Large Vision-Language Models via Monitoring Hidden States

Reference 40

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:55.207937Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3f97aa5c-0fa7-46fe-bc60-1a31d2ea3caa · outbound

This paper cites an unresolved cited work.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Unresolved cited work

Reference 41

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:55.313505Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5518f8a9-0b0d-4239-aff8-99f5d3fdfdcb · outbound

This paper cites They Infer What You Meant: Models Represent Communicative Intent More Reliably Than They Act On It.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak They Infer What You Meant: Models Represent Communicative Intent More Reliably Than They Act On It

Reference 42

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:55.454000Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 074e01f3-5f73-4ca1-bd9b-859ab30fa036 · outbound

This paper cites Prefill-level Jailbreak: A Black-Box Risk Analysis of Large Language Models.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Prefill-level Jailbreak: A Black-Box Risk Analysis of Large Language Models

Reference 43

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:55.577560Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation acd5e376-82aa-4604-a0d9-afa751ed6fa2 · outbound

This paper cites Locating and Editing Factual Associations in GPT.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Locating and Editing Factual Associations in GPT

Reference 44

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:55.735960Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 38200510-10ac-437d-8e64-518b75d22ad4 · outbound

This paper cites Pytorch: An imperative style, high-performance deep learning library.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Pytorch: An imperative style, high-performance deep learning library

Reference 45

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:55.952145Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 60cdd544-a2da-4832-b3a9-8698c54cf68f · outbound

This paper cites Scikit-learn: Machine learning in python.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Scikit-learn: Machine learning in python

Reference 46

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:56.114183Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 54e89284-a98d-4c11-8add-40007f5d5d99 · outbound

This paper cites Safety Alignment Should Be Made More Than Just a Few Tokens Deep.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Safety Alignment Should Be Made More Than Just a Few Tokens Deep

Reference 47

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:56.251349Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7b160040-6e0a-4f45-9244-f75d8d1a9334 · outbound

This paper cites Qwen2.5 Technical Report.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Qwen2.5 Technical Report

Reference 48

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:56.355016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 77d348d4-989c-415f-8233-89b4325eacb5 · outbound

This paper cites XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models

Reference 49

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:56.465152Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2cf8436d-1764-4e0e-9e72-e5d0acc30d5c · outbound

This paper cites Hashimoto.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Hashimoto

Reference 50

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:56.607129Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b2c6376f-b701-4b36-9a19-1227d8562983 · outbound

This paper cites Jailbroken: How Does LLM Safety Training Fail?.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Jailbroken: How Does LLM Safety Training Fail?

Reference 51

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:56.670122Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a5b8f13a-3dc1-4bff-874e-9de5715323a7 · outbound

This paper cites Transformers: State-of-the-art natural language processing.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Transformers: State-of-the-art natural language processing

Reference 52

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:56.786859Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1d0dae8b-2032-4f5f-bc6f-a33cfe7a3eaa · outbound

This paper cites Zhang, A.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Zhang, A

Reference 53

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:56.868524Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 01477f78-7530-4dde-b7eb-922974ec845e · outbound

This paper cites Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems

Reference 54

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:56.999697Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 027ce574-3505-4a46-b27a-6034f87b5816 · outbound

This paper cites JBShield: Defending Large Language Models from Jailbreak Attacks through Activated Concept Analysis and Manipulation.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak JBShield: Defending Large Language Models from Jailbreak Attacks through Activated Concept Analysis and Manipulation

Reference 55

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:57.107494Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cb764111-1690-49cf-b86d-588f9ad8906e · outbound

This paper cites Zhang, T.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Zhang, T

Reference 56

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:57.210550Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation dc6ee59f-46ba-45b8-8bcd-47b207960a76 · outbound

This paper cites Advprefix: An objective for nuanced llm jailbreaks.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Advprefix: An objective for nuanced llm jailbreaks

Reference 57

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:57.331387Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 10b8cbda-00c7-4fde-be39-84d18798c58b · outbound

This paper cites Representation Engineering: A Top-Down Approach to AI Transparency.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Representation Engineering: A Top-Down Approach to AI Transparency

Reference 58

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:57.467173Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3ab08afe-e139-494c-8ada-84ea7fce3676 · outbound

This paper cites Universal and Transferable Adversarial Attacks on Aligned Language Models.

Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak Universal and Transferable Adversarial Attacks on Aligned Language Models

Reference 59

Resolution: unresolved; no resolver link, observed 2026-08-02T06:30:57.585806Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

No inbound Pith citation observations are available.