English
Related papers

Related papers: Backtracking Improves Generation Safety

200 papers

Large Language Models (LLMs) can acquire deceptive behaviors through backdoor attacks, where the model executes prohibited actions whenever secret triggers appear in the input. Existing safety training methods largely fail to address this…

Cryptography and Security · Computer Science 2025-10-08 Guangyu Shen , Siyuan Cheng , Xiangzhe Xu , Yuan Zhou , Hanxi Guo , Zhuo Zhang , Xiangyu Zhang

Language model detoxification aims to minimize the risk of generating offensive or harmful content in pretrained language models (PLMs) for safer deployment. Existing methods can be roughly categorized as finetuning-based and…

Computation and Language · Computer Science 2023-10-17 Chak Tou Leong , Yi Cheng , Jiashuo Wang , Jian Wang , Wenjie Li

Despite the impressive performance of general-purpose large language models (LLMs), they often require fine-tuning or post-training to excel at specific tasks. For instance, large reasoning models (LRMs), such as the DeepSeek-R1 series,…

Computation and Language · Computer Science 2026-04-02 Mingjie Li , Wai Man Si , Michael Backes , Yang Zhang , Yisen Wang

Large language models (LLMs) have exhibited remarkable versatility and adaptability, while their widespread adoption across various applications also raises critical safety concerns. This paper focuses on the impact of backdoored LLMs.…

Computation and Language · Computer Science 2025-09-03 Jiyang Qiu , Xinbei Ma , Zhuosheng Zhang , Hai Zhao , Yun Li , Qianren Wang

Fine-tuning large language models (LLMs) for downstream tasks often leads to catastrophic forgetting, notably degrading the safety of originally aligned models. While some existing methods attempt to restore safety by incorporating…

Computation and Language · Computer Science 2025-08-29 Hua Farn , Hsuan Su , Shachi H Kumar , Saurav Sahay , Shang-Tse Chen , Hung-yi Lee

Textual backdoor attacks present a substantial security risk to Large Language Models (LLM). It embeds carefully chosen triggers into a victim model at the training stage, and makes the model erroneously predict inputs containing the same…

Computation and Language · Computer Science 2024-07-08 Xinglin Li , Xianwen He , Yao Li , Minhao Cheng

As large language models (LLMs) continue to advance in capabilities, ensuring their safety against jailbreak attacks remains a critical challenge. In this paper, we introduce a novel safety alignment approach called Answer-Then-Check, which…

Machine Learning · Computer Science 2026-03-09 Chentao Cao , Xiaojun Xu , Bo Han , Hang Li

This study addresses a critical gap in safety tuning practices for Large Language Models (LLMs) by identifying and tackling a refusal position bias within safety tuning data, which compromises the models' ability to appropriately refuse…

Computation and Language · Computer Science 2025-05-26 Youliang Yuan , Wenxiang Jiao , Wenxuan Wang , Jen-tse Huang , Jiahao Xu , Tian Liang , Pinjia He , Zhaopeng Tu

Training large language models to follow instructions makes them perform better on a wide range of tasks and generally become more helpful. However, a perfectly helpful model will follow even the most malicious instructions and readily…

Computation and Language · Computer Science 2024-03-20 Federico Bianchi , Mirac Suzgun , Giuseppe Attanasio , Paul Röttger , Dan Jurafsky , Tatsunori Hashimoto , James Zou

Language models are often trained to maximize the likelihood of the next token given past tokens in the training dataset. However, during inference time, they are utilized differently, generating text sequentially and auto-regressively by…

Machine Learning · Computer Science 2025-01-22 Zhepeng Cen , Yao Liu , Siliang Zeng , Pratik Chaudhari , Huzefa Rangwala , George Karypis , Rasool Fakoor

While large language models (LLMs) exhibit significant utility across various domains, they simultaneously are susceptible to exploitation for unethical purposes, including academic misconduct and dissemination of misinformation.…

Computation and Language · Computer Science 2024-09-24 Navid Ayoobi , Lily Knab , Wen Cheng , David Pantoja , Hamidreza Alikhani , Sylvain Flamant , Jin Kim , Arjun Mukherjee

With the widespread application of Large Language Models (LLMs), their associated security issues have become increasingly prominent, severely constraining their trustworthy deployment in critical domains. This paper proposes a novel safety…

Artificial Intelligence · Computer Science 2025-11-18 Qi Li , Jianjun Xu , Pingtao Wei , Jiu Li , Peiqiang Zhao , Jiwei Shi , Xuan Zhang , Yanhui Yang , Xiaodong Hui , Peng Xu , Wenqin Shao

Although many large language models (LLMs) have been trained to refuse harmful requests, they are still vulnerable to jailbreaking attacks which rewrite the original prompt to conceal its harmful intent. In this paper, we propose a new…

Computation and Language · Computer Science 2024-06-10 Yihan Wang , Zhouxing Shi , Andrew Bai , Cho-Jui Hsieh

Backdoor attacks manipulate model predictions by inserting innocuous triggers into training and test data. We focus on more realistic and more challenging clean-label attacks where the adversarial training examples are correctly labeled.…

Machine Learning · Computer Science 2023-10-31 Wencong You , Zayd Hammoudeh , Daniel Lowd

We explore a new language model inversion problem under strict black-box, zero-shot, and limited data conditions. We propose a novel training-free framework that reconstructs prompts using only a limited number of text outputs from a…

Computation and Language · Computer Science 2025-02-18 Hanqing Li , Diego Klabjan

This study reveals a previously unexplored vulnerability in the safety alignment of Large Language Models (LLMs). Existing aligned LLMs predominantly respond to unsafe queries with refusals, which often begin with a fixed set of prefixes…

Cryptography and Security · Computer Science 2026-01-28 Yangyang Guo , Ziwei Xu , Si Liu , Zhiming Zheng , Mohan Kankanhalli

Ensuring the safety and alignment of large language models (LLMs) with human values is crucial for generating responses that are beneficial to humanity. While LLMs have the capability to identify and avoid harmful queries, they remain…

Computation and Language · Computer Science 2024-10-22 Yihua Zhou , Xiaochuan Shi

Here, we show that current LLM unlearning methods inherently reduce models' robustness, causing them to misbehave even when a single non-adversarial forget-token is present in the retain-query. Toward understanding underlying causes, we…

Computation and Language · Computer Science 2026-04-21 Dang Huu-Tien , Hoang Thanh-Tung , Anh Bui , Minh-Phuong Nguyen , Le-Minh Nguyen , Naoya Inoue

Fine-tuning safety-aligned large language models (LLMs) can substantially compromise their safety. Previous approaches require many safety samples or calibration sets, which not only incur significant computational overhead during…

Machine Learning · Computer Science 2026-01-07 Jiawen Zhang , Lipeng He , Kejia Chen , Jian Lou , Jian Liu , Xiaohu Yang , Ruoxi Jia

The trend towards large language models (LLMs) for guardrailing against undesired behaviors is increasing and has shown promise for censoring user inputs. However, increased latency, memory consumption, hosting expenses and non-structured…

Computation and Language · Computer Science 2025-04-30 James O' Neill , Santhosh Subramanian , Eric Lin , Vaikkunth Mugunthan