{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,11,15]],"date-time":"2025-11-15T07:13:24Z","timestamp":1763190804127,"version":"3.45.0"},"reference-count":52,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,6,30]],"date-time":"2025-06-30T00:00:00Z","timestamp":1751241600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,6,30]],"date-time":"2025-06-30T00:00:00Z","timestamp":1751241600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,6,30]]},"DOI":"10.1109\/ijcnn64981.2025.11227669","type":"proceedings-article","created":{"date-parts":[[2025,11,14]],"date-time":"2025-11-14T18:46:15Z","timestamp":1763145975000},"page":"1-8","source":"Crossref","is-referenced-by-count":0,"title":["Working Towards Toxic datasets for LLM Safeguarding"],"prefix":"10.1109","author":[{"given":"Liuye","family":"Guo","sequence":"first","affiliation":[{"name":"Nanjing University of Posts and Telecommunications,School of Computer Science and Technology,Nanjing,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ziyuan","family":"Wang","sequence":"additional","affiliation":[{"name":"Nanjing University of Posts and Telecommunications,School of Computer Science and Technology,Nanjing,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhipeng","family":"Wang","sequence":"additional","affiliation":[{"name":"Nanjing University,State Key Laboratory for Novel Software Technology,Nanjing,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Tieke","family":"He","sequence":"additional","affiliation":[{"name":"Nanjing University,State Key Laboratory for Novel Software Technology,Nanjing,China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"journal-title":"OpenAI, Introducing ChatGPT","year":"2024","key":"ref1"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1145\/3605943"},{"journal-title":"Contrastive preference optimization: Pushing the boundaries of llm performance in machine translation[J]","year":"2024","author":"Xu","key":"ref3"},{"journal-title":"A comprehensive survey on process-oriented automatic text summarization with exploration of llm-based methods[J]","year":"2024","author":"Jin","key":"ref4"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/3643795.3648392"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-emnlp.272"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1145\/3442188.3445924"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.findings-acl.165"},{"journal-title":"Exploiting novel GPT-4 APIs[J]","year":"2023","author":"Pelrine","key":"ref9"},{"volume-title":"Gpt-4 technical report","year":"2024","key":"ref10"},{"journal-title":"Scisafeeval: a comprehensive benchmark for safety alignment of large language models in scientific tasks[J]","year":"2024","author":"Li","key":"ref11"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.51593\/2021ca003"},{"journal-title":"The radicalization risks of GPT-3 and advanced neural language models[J]","year":"2020","author":"McGuffie","key":"ref13"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.acl-long.229"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1145\/3531146.3533229"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/SPW63631.2024.00018"},{"journal-title":"Safety assessment of chinese large language models[J]","year":"2023","author":"Sun","key":"ref17"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.findings-emnlp.301"},{"journal-title":"A survey of reinforcement learning from human feedback[J]","year":"2023","author":"Kaufmann","key":"ref19"},{"journal-title":"Mistral 7B[J]","year":"2023","author":"Jiang","key":"ref20"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-emnlp.143"},{"key":"ref22","first-page":"27730","article-title":"Training language models to follow instructions with human feedback[J]","volume":"35","author":"Ouyang","year":"2022","journal-title":"Advances in neural information processing systems"},{"journal-title":"OpenAI training and inference costs could reach 7bn for 2024, AI startup set to lose 5bn - report: Details leak about its Microsoft Azure compute cluster, datacenterdynamics","year":"2024","author":"Moss","key":"ref23"},{"journal-title":"Llama: Open and efficient foundation language models[J]","year":"2023","author":"Touvron","key":"ref24"},{"journal-title":"Emergent abilities of large language models[J]","year":"2022","author":"Wei","key":"ref25"},{"journal-title":"Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?[J]","key":"ref26"},{"key":"ref27","first-page":"24824","article-title":"Chain-of-thought prompting elicits reasoning in large language models[J]","volume":"35","author":"Wei","year":"2022","journal-title":"Advances in neural information processing systems"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1145\/3637528.3671458"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-emnlp.243"},{"journal-title":"Meta, Introducing Meta Llama 3: The most capable openly available LLM to date","year":"2024","key":"ref30"},{"key":"ref31","article-title":"More RLHF, More Trust?","author":"Li","year":"2024","journal-title":"On The Impact of Human Preference Alignment On Language Model Trustworthiness[J]"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1145\/3770749"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1145\/3571730"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-acl.847"},{"key":"ref35","article-title":"Emergent deception and emergent optimization[J]","volume":"19","author":"Steinhardt","year":"2023","journal-title":"Bounded Regret, 2023"},{"key":"ref36","first-page":"30","article-title":"Deep reinforcement learning from human preferences[J]","author":"Christiano","year":"2017","journal-title":"Advances in neural information processing systems"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1016\/j.artint.2019.103201"},{"key":"ref38","first-page":"36","article-title":"Beavertails: Towards improved safety alignment of llm via a human-preference dataset[J]","author":"Ji","year":"2024","journal-title":"Advances in Neural Information Processing Systems"},{"journal-title":"Red teaming language models to reduce harms: Methods, scaling behaviors, and lessons learned[J]","year":"2022","author":"Ganguli","key":"ref39"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.acl-long.234"},{"journal-title":"Gptfuzzer: Red teaming large language models with auto-generated jailbreak prompts[J]","year":"2023","author":"Yu","key":"ref41"},{"journal-title":"Evaluating and mitigating discrimination in language model decisions[J]","year":"2023","author":"Tamkin","key":"ref42"},{"volume-title":"OpenAI, GPT-3.5 Turbo","year":"2024","key":"ref43"},{"journal-title":"Llama 2: Open foundation and fine-tuned chat models[J]","year":"2023","author":"Touvron","key":"ref44"},{"journal-title":"Gemma: Open models based on gemini research and technology[J]","year":"2024","author":"Team","key":"ref45"},{"journal-title":"Open clone of openai\u2019s unreleased webtext dataset scraper[J]","year":"2019","author":"Peterson","key":"ref46"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.acl-long.416"},{"article-title":"DecodingTrust: A Comprehensive Assessment of Trustworthiness in GPT Models[C]","volume-title":"NeurIPS","author":"Wang","key":"ref48"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-emnlp.388"},{"journal-title":"Training a helpful and harmless assistant with reinforcement learning from human feedback[J]","year":"2022","author":"Bai","key":"ref50"},{"journal-title":"Deceiving google\u2019s perspective api built for detecting toxic comments[J]","year":"2017","author":"Hosseini","key":"ref51"},{"article-title":"Aligning ai with shared human values","year":"2020","author":"Hendrycks","key":"ref52"}],"event":{"name":"2025 International Joint Conference on Neural Networks (IJCNN)","start":{"date-parts":[[2025,6,30]]},"location":"Rome, Italy","end":{"date-parts":[[2025,7,5]]}},"container-title":["2025 International Joint Conference on Neural Networks (IJCNN)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11227166\/11227148\/11227669.pdf?arnumber=11227669","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,15]],"date-time":"2025-11-15T07:10:45Z","timestamp":1763190645000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11227669\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,30]]},"references-count":52,"URL":"https:\/\/doi.org\/10.1109\/ijcnn64981.2025.11227669","relation":{},"subject":[],"published":{"date-parts":[[2025,6,30]]}}}