English
Related papers

Related papers: DSPA: Dynamic SAE Steering for Data-Efficient Pref…

200 papers

Although LLMs have achieved significant success, their reliance on large volumes of human-annotated data has limited their potential for further scaling. In this situation, utilizing self-generated synthetic data has become crucial for…

Computation and Language · Computer Science 2026-03-17 Haoyan Yang , Khiem Le , Ting Hua , Shangqian Gao , Binfeng Xu , Zheng Tang , Jie Xu , Nitesh V. Chawla , Hongxia Jin , Vijay Srinivasan

Adapters are a parameter-efficient alternative to fine-tuning, which augment a frozen base network to learn new tasks. Yet, the inference of the adapted model is often slower than the corresponding fine-tuned model. To improve on this, we…

Computer Vision and Pattern Recognition · Computer Science 2023-02-03 Lukas Hedegaard , Aman Alok , Juby Jose , Alexandros Iosifidis

We propose a statistical adaptive procedure called SALSA for automatically scheduling the learning rate (step size) in stochastic gradient methods. SALSA first uses a smoothed stochastic line-search procedure to gradually increase the…

Machine Learning · Statistics 2020-02-26 Pengchuan Zhang , Hunter Lang , Qiang Liu , Lin Xiao

Human alignment in large language models (LLMs) is an active area of research. A recent groundbreaking work, direct preference optimization (DPO), has greatly simplified the process from past work in reinforcement learning from human…

Computation and Language · Computer Science 2025-03-10 Changyu Chen , Zichen Liu , Chao Du , Tianyu Pang , Qian Liu , Arunesh Sinha , Pradeep Varakantham , Min Lin

Attention-based Transformers have revolutionized natural language processing (NLP) and shown strong performance in computer vision (CV) tasks. However, as the input sequence varies, the computational bottlenecks in Transformer models…

Machine Learning · Computer Science 2025-12-10 Huizheng Wang , Hongbin Wang , Shaojun Wei , Yang Hu , Shouyi Yin

DPO is an effective preference optimization algorithm. However, the DPO-tuned models tend to overfit on the dispreferred samples, manifested as overly long generations lacking diversity. While recent regularization approaches have…

Computation and Language · Computer Science 2025-08-26 Chenxu Yang , Ruipeng Jia , Naibin Gu , Zheng Lin , Siyuan Chen , Chao Pang , Weichong Yin , Yu Sun , Hua Wu , Weiping Wang

Aligning large language models (LLMs) with human values is an increasingly critical step in post-training. Direct Preference Optimization (DPO) has emerged as a simple, yet effective alternative to reinforcement learning from human feedback…

Artificial Intelligence · Computer Science 2025-07-29 Yifan Wang , Runjin Chen , Bolian Li , David Cho , Yihe Deng , Ruqi Zhang , Tianlong Chen , Zhangyang Wang , Ananth Grama , Junyuan Hong

Attention steering is an important technique for controlling model focus, enabling capabilities such as prompt highlighting, where the model prioritises user-specified text. However, existing attention steering methods require explicit…

Computation and Language · Computer Science 2026-03-03 Weixian Waylon Li , Yuchen Niu , Yongxin Yang , Keshuang Li , Tiejun Ma , Shay B. Cohen

Sparse Autoencoders (SAEs) are widely employed for mechanistic interpretability and model steering. Within this context, steering is by design performed by means of decoding altered SAE intermediate representations. This procedure…

Machine Learning · Computer Science 2025-12-08 Antonio Bărbălau , Cristian Daniel Păduraru , Teodor Poncu , Alexandru Tifrea , Elena Burceanu

DPO has become a widely adopted alternative to RLHF for aligning LLMs with human preferences, eliminating the need for a separate reward model or RL loop. Recent theoretical analysis uncovers an asymmetric gradient behavior in DPO: the loss…

Computation and Language · Computer Science 2026-05-28 Shaolong Chen , Madalina Ciobanu , Qingqing Mao , Ritankar Das

The increasing demand for long-context modeling in large language models (LLMs) is bottlenecked by the quadratic complexity of the standard self-attention mechanism. The community has proposed sparse attention to mitigate this issue.…

Artificial Intelligence · Computer Science 2025-11-18 Jingze Shi , Yifan Wu , Yiran Peng , Bingheng Wu , Liangdong Wang , Guang Liu , Yuyu Luo

Direct Preference Optimization is an offline post-SFT method for aligning language models from preference pairs, with strong results in instruction following and summarization. However, DPO's sequence-level implicit reward can be brittle…

Computation and Language · Computer Science 2026-03-03 Samah Fodeh , Linhai Ma , Ganesh Puthiaraju , Srivani Talakokkul , Afshan Khan , Ashley Hagaman , Sarah R. Lowe , Aimee Kendall Roundtree

Our goal is to enable large language models (LLMs) to balance multiple human preference dimensions; such as helpfulness, safety, and verbosity, through principled and controllable alignment. Existing preference optimization methods,…

Machine Learning · Computer Science 2026-02-03 Mete Erdogan

Large-scale text-to-image foundation models have achieved remarkable visual realism, yet generating human images with correct anatomical structures remains challenging. Existing approaches enforce anatomical constraints through…

Computer Vision and Pattern Recognition · Computer Science 2026-05-26 Bao Li , Yuliang Xiu , Zhen Liu

Sparse autoencoders (SAEs) have recently emerged as a powerful tool for language model steering. Prior work has explored top-k SAE latents for steering, but we observe that many dimensions among the top-k latents capture non-semantic…

Computation and Language · Computer Science 2025-10-03 Jiaqing Xie

Most existing methods for unsupervised domain adaptation (UDA) rely on a shared network to extract domain-invariant features. However, when facing multiple source domains, optimizing such a network involves updating the parameters of the…

Computer Vision and Pattern Recognition · Computer Science 2024-05-31 Haoran Chen , Xintong Han , Zuxuan Wu , Yu-Gang Jiang

Preference optimization is a critical post-training technique used to align large language models (LLMs) with human preferences, typically by fine-tuning on ranked response pairs. While methods like Direct Preference Optimization (DPO) have…

Computation and Language · Computer Science 2025-11-12 Rhitabrat Pokharel , Yufei Tao , Ameeta Agrawal

Foundation models (FMs) are pre-trained on large-scale datasets and then fine-tuned for a specific downstream task. The most common fine-tuning method is to update pretrained weights via low-rank adaptation (LoRA). Existing initialization…

Machine Learning · Computer Science 2025-10-21 Fabian Paischer , Lukas Hauzenberger , Thomas Schmied , Benedikt Alkin , Marc Peter Deisenroth , Sepp Hochreiter

Direction-of-arrival (DOA) estimation is an important task in microphone array processing and many downstream applications. The steered response power with phase transform (SRP-PHAT) method has been widely adopted for DOA estimation in…

Audio and Speech Processing · Electrical Eng. & Systems 2026-04-29 Ming Huang , Shuting Xu , Leying Yang , Huanzhang Hu , Yujie Zhang , Jiang Wang , Yu Liu , Hao Zhao , He Kong

We propose an efficient fine-tuning method for time series foundation models, termed TRACE: Time Series Parameter Efficient Fine-tuning. While pretrained time series foundation models are gaining popularity, they face the following…

Machine Learning · Computer Science 2025-11-24 Yuze Li , Wei Zhu