SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models | Paper Citation Record

Observation aed38e89-35de-4a6d-921c-1a46b2b99efa · outbound

This paper cites Improving language understanding by generative pre-training2018.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Improving language understanding by generative pre-training2018

Reference 1

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:04.904571Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 1ef3f846-4409-4170-a3c4-4c062658fd0d · outbound

This paper cites GPT-4 Technical Report.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models GPT-4 Technical Report

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:51.894501Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d3dd119b-d654-4d6b-b168-1bdb893da6c2 · outbound

This paper cites DeepSeek LLM: Scaling Open-Source Language Models with Longtermism.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models DeepSeek LLM: Scaling Open-Source Language Models with Longtermism

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:51.997871Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 475ed272-df6e-433f-8369-58ebefe17c1a · outbound

This paper cites DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:52.145353Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2f51017b-8333-451b-9117-773d2e2a4d40 · outbound

This paper cites PAL: Proxy-Guided Black-Box Attack on Large Language Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models PAL: Proxy-Guided Black-Box Attack on Large Language Models

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:52.222527Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 397bd4f2-0c8f-40b2-bb8e-1e50f9542ab5 · outbound

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

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Universal and Transferable Adversarial Attacks on Aligned Language Models

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:52.358358Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cc1567fc-3ba2-4fc9-9c42-1ccde88ab3d8 · outbound

This paper cites Monkey: Image Resolution and Text Label Are Important Things for Large Multi-modal Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Monkey: Image Resolution and Text Label Are Important Things for Large Multi-modal Models

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:52.542586Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3cf6324c-550c-4d8a-8cbb-105e3cb79ade · outbound

This paper cites A survey of backdoor attacks and defenses on large language models: Implications for security measures.Authorea Preprints2024.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models A survey of backdoor attacks and defenses on large language models: Implications for security measures.Authorea Preprints2024

Reference 8

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:04.636289Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation b8189ea6-8014-43c2-9a13-f58b06787a87 · outbound

This paper cites Training language models to follow instructions with human feedback.NeurIPS2022,35, 27730–27744.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Training language models to follow instructions with human feedback.NeurIPS2022,35, 27730–27744

Reference 9

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:04.274233Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation c6aa9a52-c618-443e-b245-10472b891979 · outbound

This paper cites Defending Large Language Models Against Jailbreak Attacks Through Chain of Thought Prompting.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Defending Large Language Models Against Jailbreak Attacks Through Chain of Thought Prompting

Reference 10

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:04.119138Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation ad0c9440-6356-4961-8102-c23bd7bf22f1 · outbound

This paper cites Many-shot jailbreaking.NeurIPS2024,37, 129696–129742.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Many-shot jailbreaking.NeurIPS2024,37, 129696–129742

Reference 11

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:03.956256Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 1107255d-f51b-401a-841e-7976b9a6d564 · outbound

This paper cites Pandora: Jailbreak GPTs by Retrieval Augmented Generation Poisoning.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Pandora: Jailbreak GPTs by Retrieval Augmented Generation Poisoning

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:53.234749Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2ec64b4f-08f6-42c6-a595-c514a2f21e81 · outbound

This paper cites Multi-step Jailbreaking Privacy Attacks on ChatGPT.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Multi-step Jailbreaking Privacy Attacks on ChatGPT

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:53.545039Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5b09016f-1598-49f5-8296-9045b92df4b2 · outbound

This paper cites Adversarial Demonstration Attacks on Large Language Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Adversarial Demonstration Attacks on Large Language Models

Reference 14

Resolution: malformed identifier; no resolver link, observed 2026-08-06T23:20:53.674318Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation acd11a48-ed3f-4c4d-be94-32473105ddd4 · outbound

This paper cites Jailbreak and Guard Aligned Language Models with Only Few In-Context Demonstrations.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Jailbreak and Guard Aligned Language Models with Only Few In-Context Demonstrations

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:53.797609Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8c70371a-3c8c-4c98-8d98-02c6c726e70b · outbound

This paper cites Improved few-shot jailbreaking can circumvent aligned language models and their defenses.NeurIPS2024,37, 32856–32887.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Improved few-shot jailbreaking can circumvent aligned language models and their defenses.NeurIPS2024,37, 32856–32887

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:03.744360Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation c19d5bb5-90b4-41b0-a6fe-a093ffc61213 · outbound

This paper cites Play Guessing Game with LLM: Indirect Jailbreak Attack with Implicit Clues.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Play Guessing Game with LLM: Indirect Jailbreak Attack with Implicit Clues

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:53.974751Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation abd74381-d8d7-4181-8228-b93e0156df97 · outbound

This paper cites Artprompt: Ascii art-based jailbreak attacks against aligned llms.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Artprompt: Ascii art-based jailbreak attacks against aligned llms

Reference 18

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:03.585017Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 7934abf5-bb74-4271-a43f-dfa0858bc75c · outbound

This paper cites Semantic Mirror Jailbreak: Genetic Algorithm Based Jailbreak Prompts Against Open-source LLMs.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Semantic Mirror Jailbreak: Genetic Algorithm Based Jailbreak Prompts Against Open-source LLMs

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:54.194806Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 62c6b020-c4b5-4cfb-ba2b-4431d76ecd1d · outbound

This paper cites Understanding and Enhancing the Transferability of Jailbreaking Attacks.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Understanding and Enhancing the Transferability of Jailbreaking Attacks

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:54.264785Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 04bf703d-e726-4859-8139-7bd43ff41a54 · outbound

This paper cites AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Models

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:54.385546Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f737caa1-dd19-4063-b49f-336cf5337eaf · outbound

This paper cites Flipattack: Jailbreak llms via flipping.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Flipattack: Jailbreak llms via flipping

Reference 22

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:03.365496Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation b80279ff-90d9-451b-bac6-6d4c9d7bfbff · outbound

This paper cites All in how you ask for it: Simple black-box method for jailbreak attacks.Applied Sciences2024,14, 3558.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models All in how you ask for it: Simple black-box method for jailbreak attacks.Applied Sciences2024,14, 3558

Reference 23

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:03.130334Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 66031097-22d5-4869-95db-3bd8274d5547 · outbound

This paper cites Emoji Attack: Enhancing Jailbreak Attacks Against Judge LLM Detection.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Emoji Attack: Enhancing Jailbreak Attacks Against Judge LLM Detection

Reference 24

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:02.924635Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 572595dc-71e9-4f77-ab83-eafeb7994cbd · outbound

This paper cites The Dark Side of Trust: Authority Citation-Driven Jailbreak Attacks on Large Language Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models The Dark Side of Trust: Authority Citation-Driven Jailbreak Attacks on Large Language Models

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:54.845285Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 80a5b13b-f9b0-4d0d-8bd2-195e885fbab4 · outbound

This paper cites GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:54.935278Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fe8b4cf5-c9ed-4ddc-9b02-460cffe25f72 · outbound

This paper cites Gpt-4 is too smart to be safe: Stealthy chat with llms via cipher2024.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Gpt-4 is too smart to be safe: Stealthy chat with llms via cipher2024

Reference 27

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:02.700975Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 54a66195-b915-43e4-9959-84fd95786d5c · outbound

This paper cites Explore, Establish, Exploit: Red Teaming Language Models from Scratch.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Explore, Establish, Exploit: Red Teaming Language Models from Scratch

Reference 28

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:55.144189Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c04a6230-7c1e-48d7-a277-2891647b17b5 · outbound

This paper cites Jailbreaking black box large language models in twenty queries.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Jailbreaking black box large language models in twenty queries

Reference 29

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:02.554468Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 4e48964a-e6cf-40b9-95be-f7f0537b1c4b · outbound

This paper cites MasterKey: Automated Jailbreak Across Multiple Large Language Model Chatbots.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models MasterKey: Automated Jailbreak Across Multiple Large Language Model Chatbots

Reference 30

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:55.350714Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 55094c79-bba4-4d98-b270-58aefc9fa2a9 · outbound

This paper cites MART: Improving LLM Safety with Multi-round Automatic Red-Teaming.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models MART: Improving LLM Safety with Multi-round Automatic Red-Teaming

Reference 31

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:02.386229Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 4c262b5e-e8cb-4176-8743-7c89d623960d · outbound

This paper cites Guard: Role-playing to generate natural-language jailbreakings to test guideline adherence of large language models.arXiv preprint arXiv:2402.032992024.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Guard: Role-playing to generate natural-language jailbreakings to test guideline adherence of large language models.arXiv preprint arXiv:2402.032992024

Reference 32

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:55.484326Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1fc88740-52ae-418d-8b91-0abd30c85890 · outbound

This paper cites Goal-Oriented Prompt Attack and Safety Evaluation for LLMs.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Goal-Oriented Prompt Attack and Safety Evaluation for LLMs

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:55.602649Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4ed0e8b6-e09e-46ab-8341-c7523c28610e · outbound

This paper cites Tree of attacks: Jailbreaking black-box llms automatically.NeurIPS2024,37, 61065–61105.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Tree of attacks: Jailbreaking black-box llms automatically.NeurIPS2024,37, 61065–61105

Reference 34

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:02.212017Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 6d906230-1a4e-4b3f-b9a8-854bff233fbf · outbound

This paper cites Scalable and Transferable Black-Box Jailbreaks for Language Models via Persona Modulation.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Scalable and Transferable Black-Box Jailbreaks for Language Models via Persona Modulation

Reference 35

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:55.859123Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 207ab80e-bb35-4fe7-b1ba-5e6a8bd3c6af · outbound

This paper cites Evil Geniuses: Delving into the Safety of LLM-based Agents.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Evil Geniuses: Delving into the Safety of LLM-based Agents

Reference 36

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:55.953808Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 898e97be-5c05-4f88-9ce5-c472c57af4b6 · outbound

This paper cites How johnny can persuade llms to jailbreak them: Rethinking persuasion to challenge ai safety by humanizing llms.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models How johnny can persuade llms to jailbreak them: Rethinking persuasion to challenge ai safety by humanizing llms

Reference 37

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:02.015166Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 09cebd2c-7789-45e8-89f2-5eb8bf3a0b10 · outbound

This paper cites Attacking Large Language Models with Projected Gradient Descent.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Attacking Large Language Models with Projected Gradient Descent

Reference 38

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:56.204745Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a469243d-98ab-448f-a553-d9a672b4148f · outbound

This paper cites Query-based adversarial prompt generation.NeurIPS2024, 37, 128260–128279.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Query-based adversarial prompt generation.NeurIPS2024, 37, 128260–128279

Reference 39

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:01.842246Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation d2bfecdd-1356-4824-8494-fac5065e7174 · outbound

This paper cites Improved Techniques for Optimization-Based Jailbreaking on Large Language Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Improved Techniques for Optimization-Based Jailbreaking on Large Language Models

Reference 40

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:56.482120Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2309b2d9-7282-4885-9774-dc260c466e86 · outbound

This paper cites Iterative self-tuning llms for enhanced jailbreaking capabilities.arXiv preprint arXiv:2410.184692024.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Iterative self-tuning llms for enhanced jailbreaking capabilities.arXiv preprint arXiv:2410.184692024

Reference 41

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:56.584927Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 980df864-1235-45f1-a729-c8df39e53787 · outbound

This paper cites From noise to clarity: Unraveling the adversarial suffix of large language model attacks via translation of text embeddings.CoRR2024.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models From noise to clarity: Unraveling the adversarial suffix of large language model attacks via translation of text embeddings.CoRR2024

Reference 42

Resolution: malformed identifier; raw_fallback, observed 2026-08-06T23:21:01.748411Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 0f629e3d-fe27-4fae-93c7-e2ab5756f978 · outbound

This paper cites Guiding not Forcing: Enhancing the Transferability of Jailbreaking Attacks on LLMs via Removing Superfluous Constraints.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Guiding not Forcing: Enhancing the Transferability of Jailbreaking Attacks on LLMs via Removing Superfluous Constraints

Reference 43

Resolution: verified exact; local_arxiv, observed 2026-08-06T23:21:00.004770Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 73d991a7-641e-443f-9e8e-375b9bf881d5 · outbound

This paper cites AutoDAN: Interpretable Gradient-Based Adversarial Attacks on Large Language Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models AutoDAN: Interpretable Gradient-Based Adversarial Attacks on Large Language Models

Reference 44

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:56.924751Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f2de84f1-870e-469e-b6f7-00b81abe62f6 · outbound

This paper cites Analyzing the Inherent Response Tendency of LLMs: Real-World Instructions-Driven Jailbreak.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Analyzing the Inherent Response Tendency of LLMs: Real-World Instructions-Driven Jailbreak

Reference 45

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:57.014753Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d6c92b03-eb4c-412c-a676-26de960e4f10 · outbound

This paper cites COLD-Attack: Jailbreaking LLMs with Stealthiness and Controllability.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models COLD-Attack: Jailbreaking LLMs with Stealthiness and Controllability

Reference 46

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:01.592880Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 4340b1d7-49c0-436b-84dc-94c912fa73fd · outbound

This paper cites DROJ: A Prompt-Driven Attack against Large Language Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models DROJ: A Prompt-Driven Attack against Large Language Models

Reference 47

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:57.184523Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 31f963ff-fdbe-4b35-b7e8-2b9f30b76df0 · outbound

This paper cites Catastrophic jailbreak of open-source llms via exploiting generation.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Catastrophic jailbreak of open-source llms via exploiting generation

Reference 48

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:01.378863Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation dbf5d646-15db-44cb-995c-4f51ac048dd8 · outbound

This paper cites Make Them Spill the Beans! Coercive Knowledge Extraction from (Production) LLMs.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Make Them Spill the Beans! Coercive Knowledge Extraction from (Production) LLMs

Reference 49

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:57.420571Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 31368c05-277e-4bbf-b0e3-6e2c92dff6f8 · outbound

This paper cites Weak-to-Strong Jailbreaking on Large Language Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Weak-to-Strong Jailbreaking on Large Language Models

Reference 50

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:57.504508Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 42a48662-4628-4735-83ee-9602c01ffcd3 · outbound

This paper cites Don't Say No: Jailbreaking LLM by Suppressing Refusal.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Don't Say No: Jailbreaking LLM by Suppressing Refusal

Reference 51

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:57.656839Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 42f69fb9-0057-4e6d-88e6-42a3df17cf2d · outbound

This paper cites Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!

Reference 52

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:57.744842Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8423a4ca-4720-4e34-9c9a-ca8ba783decd · outbound

This paper cites Shadow Alignment: The Ease of Subverting Safely-Aligned Language Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Shadow Alignment: The Ease of Subverting Safely-Aligned Language Models

Reference 53

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:57.826882Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ab1d7d35-a32d-4fb2-9219-90097f0a8d86 · outbound

This paper cites Removing RLHF Protections in GPT-4 via Fine-Tuning.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Removing RLHF Protections in GPT-4 via Fine-Tuning

Reference 54

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:57.988075Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 685c4854-50a5-45aa-9108-651a18eea961 · outbound

This paper cites HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal

Reference 55

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:58.119677Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8ace8b2b-0596-49af-b719-c484030b59f7 · outbound

This paper cites EasyJailbreak: A Unified Framework for Jailbreaking Large Language Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models EasyJailbreak: A Unified Framework for Jailbreaking Large Language Models

Reference 56

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:58.217814Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d8e38e2c-5701-411c-be1a-3df3c7fb0434 · outbound

This paper cites Playing Language Game with LLMs Leads to Jailbreaking.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Playing Language Game with LLMs Leads to Jailbreaking

Reference 57

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:58.265798Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8453da1f-f303-4dfc-a528-6b3a81731ada · outbound

This paper cites do anything now.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models do anything now

Reference 58

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:01.234729Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 9d70e8b5-b0ee-4355-816b-25171df3f631 · outbound

This paper cites Chain-of-Lure: A Synthetic Narrative-Driven Approach to Compromise Large Language Models.arXiv preprint arXiv:2505.175192025.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Chain-of-Lure: A Synthetic Narrative-Driven Approach to Compromise Large Language Models.arXiv preprint arXiv:2505.175192025

Reference 59

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:58.396202Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9cfac2d0-946a-46f1-b1ce-98acfb8deaf1 · outbound

This paper cites Reasoning-Augmented Conversation for Multi-Turn Jailbreak Attacks on Large Language Models.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Reasoning-Augmented Conversation for Multi-Turn Jailbreak Attacks on Large Language Models

Reference 60

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:58.435774Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 12e3a4ea-66ab-4847-8f75-b609b11b7db2 · outbound

This paper cites Amplified Vulnerabilities: Structured Jailbreak Attacks on LLM-based Multi-Agent Debate.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Amplified Vulnerabilities: Structured Jailbreak Attacks on LLM-based Multi-Agent Debate

Reference 61

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:58.553566Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a5edf243-33c2-4ba1-a47c-b019af836f53 · outbound

This paper cites H-CoT: Hijacking the Chain-of-Thought Safety Reasoning Mechanism to Jailbreak Large Reasoning Models, Including OpenAI o1/o3, DeepSeek-R1, and Gemini 2.0 Flash Thinking.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models H-CoT: Hijacking the Chain-of-Thought Safety Reasoning Mechanism to Jailbreak Large Reasoning Models, Including OpenAI o1/o3, DeepSeek-R1, and Gemini 2.0 Flash Thinking

Reference 62

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:58.633864Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5da767d8-d948-43cd-ae62-0c59201e73a0 · outbound

This paper cites Advancing Jailbreak Strategies: A Hybrid Approach to Exploiting LLM Vulnerabilities and Bypassing Modern Defenses.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Advancing Jailbreak Strategies: A Hybrid Approach to Exploiting LLM Vulnerabilities and Bypassing Modern Defenses

Reference 63

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:58.767251Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 780df3e5-5c8b-4e64-9949-794827c298c4 · outbound

This paper cites Gradient cuff: Detecting jailbreak attacks on large language models by exploring refusal loss landscapes.NeurIPS2024,37, 126265–126296.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Gradient cuff: Detecting jailbreak attacks on large language models by exploring refusal loss landscapes.NeurIPS2024,37, 126265–126296

Reference 64

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:01.066495Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 4b26e0de-d342-4c23-b32a-554f30637313 · outbound

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

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models JBShield: Defending Large Language Models from Jailbreak Attacks through Activated Concept Analysis and Manipulation

Reference 65

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:00.942413Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation acc9b9a0-018d-4138-8dfa-666ea874745b · outbound

This paper cites The hidden risks of large reasoning models: A safety assessment of r1.arXiv preprint arXiv:2502.126592025.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models The hidden risks of large reasoning models: A safety assessment of r1.arXiv preprint arXiv:2502.126592025

Reference 66

Resolution: unresolved; no resolver link, observed 2026-08-06T23:20:59.040436Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8625a784-0c3b-4c2e-903f-810fca884fe2 · outbound

This paper cites Scaling Trends in Language Model Robustness.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Scaling Trends in Language Model Robustness

Reference 67

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T23:21:00.662149Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 9d2998c7-079e-4b37-8077-327cbdf5afaf · outbound

This paper cites Scaling Behavior of Machine Translation with Large Language Models under Prompt Injection Attacks.

SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models Scaling Behavior of Machine Translation with Large Language Models under Prompt Injection Attacks

Reference 68

Resolution: malformed identifier; local_arxiv, observed 2026-08-06T23:20:59.349100Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Coverage vector

Reference resolution

Outbound references

Pith citing papers