English
Related papers

Related papers: Consistency Training Helps Stop Sycophancy and Jai…

200 papers

Supervised Fine-Tuning (SFT) is the standard paradigm for domain adaptation, yet it frequently incurs the cost of catastrophic forgetting. In sharp contrast, on-policy Reinforcement Learning (RL) effectively preserves general capabilities.…

Machine Learning · Computer Science 2026-01-06 Muxi Diao , Lele Yang , Wuxuan Gong , Yutong Zhang , Zhonghao Yan , Yufei Han , Kongming Liang , Weiran Xu , Zhanyu Ma

Recent advances of Reinforcement Learning (RL) have highlighted its potential in complex reasoning tasks, yet effective training often relies on external supervision, which limits the broader applicability. In this work, we propose a novel…

Artificial Intelligence · Computer Science 2025-06-11 Kongcheng Zhang , Qi Yao , Shunyu Liu , Yingjie Wang , Baisheng Lai , Jieping Ye , Mingli Song , Dacheng Tao

Large Language Models (LLMs) are commonly evaluated for robustness against paraphrased or semantically equivalent jailbreak prompts, yet little attention has been paid to linguistic variation as an attack surface. In this work, we…

Computation and Language · Computer Science 2025-11-14 Srikant Panda , Avinash Rai

Aligning Large Language Models (LLMs) with human preferences is crucial for their deployment in real-world applications. Recent advancements in Self-Rewarding Language Models suggest that an LLM can use its internal reward models (such as…

Artificial Intelligence · Computer Science 2025-02-14 Xin Zhou , Yiwen Guo , Ruotian Ma , Tao Gui , Qi Zhang , Xuanjing Huang

Many publicly available language models have been safety tuned to reduce the likelihood of toxic or liability-inducing text. To redteam or jailbreak these models for compliance with toxic requests, users and security analysts have developed…

Computation and Language · Computer Science 2024-10-02 T. Ben Thompson , Michael Sklar

Continual Instruction Tuning (CIT) is adopted to continually instruct Large Models to follow human intent data by data. It is observed that existing gradient update would heavily destroy the performance on previous datasets during CIT…

Machine Learning · Computer Science 2025-12-15 Jingyang Qiao , Zhizhong Zhang , Xin Tan , Yanyun Qu , Shouhong Ding , Yuan Xie

Large language models (LLMs) can be dishonest when reporting on their actions and beliefs -- for example, they may overstate their confidence in factual claims or cover up evidence of covert actions. Such dishonesty may arise due to the…

Machine Learning · Computer Science 2025-12-24 Manas Joglekar , Jeremy Chen , Gabriel Wu , Jason Yosinski , Jasmine Wang , Boaz Barak , Amelia Glaese

The increasing integration of Large Language Models (LLMs) into decision-making frameworks has exposed significant vulnerabilities to social compliance, specifically sycophancy and conformity. However, a critical research gap exists…

Computers and Society · Computer Science 2026-01-21 Long Zhang , Wei-neng Chen

Modern language models fail a fundamental requirement of trustworthy intelligence: knowing when not to answer. Despite achieving impressive accuracy on benchmarks, these models produce confident hallucinations, even when wrong answers carry…

Machine Learning · Computer Science 2025-11-25 Mohamad Amin Mohamadi , Tianhao Wang , Zhiyuan Li

Longitudinal NLP tasks require reasoning over temporally ordered text to detect persistence and change in human behavior and opinions. However, in-context learning with large language models struggles on tasks where models must integrate…

Computation and Language · Computer Science 2026-04-21 Iqra Ali , Talia Tseriotou , Mahmud Elahi Akhter , Yuxiang Zhou , Maria Liakata

Large language models (LLMs) have shown success in many natural language processing tasks. Despite rigorous safety alignment processes, supposedly safety-aligned LLMs like Llama 2 and Claude 2 are still susceptible to jailbreaks, leading to…

Computation and Language · Computer Science 2024-05-15 Ziyang Zhang , Qizhen Zhang , Jakob Foerster

Standard attention-based transformers are known to exhibit instability under learning rate overspecification during training, particularly at high learning rates. While various methods have been proposed to improve resilience to such…

Machine Learning · Computer Science 2026-02-02 Shyam Venkatasubramanian , Sean Moushegian , Michael Lin , Mir Park , Ankit Singhal , Connor Lee

In the context of continual learning, prototypes-as representative class embeddings-offer advantages in memory conservation and the mitigation of catastrophic forgetting. However, challenges related to semantic drift and prototype…

Machine Learning · Computer Science 2023-11-14 Zhuowei Li , Long Zhao , Zizhao Zhang , Han Zhang , Di Liu , Ting Liu , Dimitris N. Metaxas

Instruction-tuned large language models produce helpful, structured responses, but how robust is this helpfulness under trivial constraints? We show that simple lexical constraints (banning a single punctuation character or common word)…

Computation and Language · Computer Science 2026-04-28 Erfan Baghaei Potraghloo , Seyedarmin Azizi , Souvik Kundu , Massoud Pedram

Large Language Models (LLMs) for public use require continuous pre-training to remain up-to-date with the latest data. The models also need to be fine-tuned with specific instructions to maintain their ability to follow instructions…

Computation and Language · Computer Science 2024-10-15 Ishan Jindal , Chandana Badrinath , Pranjal Bharti , Lakkidi Vinay , Sachin Dev Sharma

Most sequential recommendation models capture the features of consecutive items in a user-item interaction history. Though effective, their representation expressiveness is still hindered by the sparse learning signals. As a result, the…

Information Retrieval · Computer Science 2021-06-30 Xu Yuan , Hongshen Chen , Yonghao Song , Xiaofang Zhao , Zhuoye Ding , Zhen He , Bo Long

Reducing inconsistencies in the behavior of different versions of an AI system can be as important in practice as reducing its overall error. In image classification, sample-wise inconsistencies appear as "negative flips": A new model…

Computer Vision and Pattern Recognition · Computer Science 2021-05-19 Sijie Yan , Yuanjun Xiong , Kaustav Kundu , Shuo Yang , Siqi Deng , Meng Wang , Wei Xia , Stefano Soatto

Prompt-based continual learning methods fine-tune only a small set of additional learnable parameters while keeping the pre-trained model's parameters frozen. It enables efficient adaptation to new tasks while mitigating the risk of…

Computer Vision and Pattern Recognition · Computer Science 2026-05-11 Shengqin Jiang , Tianqi Kong , Yuankai Qi , Haokui Zhang , Lina Yao , Quan Z. Sheng , Qingshan Liu , Ming-Hsuan Yang

Large visual language models (LVLMs) have demonstrated excellent instruction-following capabilities, yet remain vulnerable to stealthy backdoor attacks when finetuned using contaminated data. Existing backdoor defense techniques are usually…

Cryptography and Security · Computer Science 2025-06-09 Yuan Xun , Siyuan Liang , Xiaojun Jia , Xinwei Liu , Xiaochun Cao

Consistency models are a new family of generative models capable of producing high-quality samples in either a single step or multiple steps. Recently, consistency models have demonstrated impressive performance, achieving results on par…

Computer Vision and Pattern Recognition · Computer Science 2025-03-26 Quan Dao , Khanh Doan , Di Liu , Trung Le , Dimitris Metaxas