English
Related papers

Related papers: The Distillation Game: Adaptive Attacks & Efficien…

200 papers

Training-time privileged information (PI) can enable language models to succeed on tasks they would otherwise fail, making it a powerful tool for reinforcement learning in hard, long-horizon settings. However, transferring capabilities…

Machine Learning · Computer Science 2026-02-17 Emiliano Penaloza , Dheeraj Vattikonda , Nicolas Gontier , Alexandre Lacoste , Laurent Charlin , Massimo Caccia

The quality of open-weight language models has dramatically improved in recent years. Sharing weights greatly facilitates model adoption by enabling their use across diverse hardware and software platforms. They also allow for more open…

Machine Learning · Computer Science 2026-05-12 Keitaro Sakamoto , Pierre Ablin , Federico Danieli , Marco Cuturi

Traditionally, distillation has been used to train a student model to emulate the input/output functionality of a teacher. A more useful goal than emulation, yet under-explored, is for the student to learn feature representations that…

Computer Vision and Pattern Recognition · Computer Science 2021-07-19 Zhizhong Li , Avinash Ravichandran , Charless Fowlkes , Marzia Polito , Rahul Bhotika , Stefano Soatto

On-policy distillation (OPD) is increasingly used in LLM post-training because it can leverage a teacher model to provide dense supervision on student rollouts. The standard implementation, however, usually reduces distribution matching to…

Machine Learning · Computer Science 2026-04-28 Yuqian Fu , Haohuan Huang , Kaiwen Jiang , Jiacai Liu , Zhuo Jiang , Yuanheng Zhu , Dongbin Zhao

Unsupervised video segmentation is a challenging computer vision task, especially due to the lack of supervisory signals coupled with the complexity of visual scenes. To overcome this challenge, state-of-the-art models based on slot…

Computer Vision and Pattern Recognition · Computer Science 2025-11-19 Diana-Nicoleta Grigore , Neelu Madan , Andreas Mogelmose , Thomas B. Moeslund , Radu Tudor Ionescu

Knowledge distillation generally assumes a strong-to-weak relationship where stronger teachers yield better students. In this work, we examine this assumption about distillation in large language model pretraining. By varying architecture…

Machine Learning · Computer Science 2026-05-25 Taiming Lu , Zhuang Liu

The results of a learning process depend on the input data. There are cases in which an adversary can strategically tamper with the input data to affect the outcome of the learning process. While some datasets are difficult to attack, many…

Cryptography and Security · Computer Science 2019-04-02 Eitan Farchi , Onn Shehory , Guy Barash

Direct Preference Optimization (DPO) is a powerful paradigm to align language models with human preferences using pairwise comparisons. However, its binary win-or-loss supervision often proves insufficient for training small models with…

Computation and Language · Computer Science 2025-09-23 Minchan Kwon , Junwon Ko , Kangil Kim , Junmo Kim

Feature distillation is an effective way to improve the performance for a smaller student model, which has fewer parameters and lower computation cost compared to the larger teacher model. Unfortunately, there is a common obstacle - the gap…

Computer Vision and Pattern Recognition · Computer Science 2020-10-14 Kaiyu Yue , Jiangfan Deng , Feng Zhou

Recent progress in accelerating text-to-image diffusion models enables high-fidelity synthesis within a single denoising step. However, customizing the fast one-step models remains challenging, as existing methods consistently fail to…

Computer Vision and Pattern Recognition · Computer Science 2026-04-13 Yixiong Yang , Tao Wu , Senmao Li , Shiqi Yang , Yaxing Wang , Joost van de Weijer , Kai Wang

The performance of autoregressive models on natural language generation tasks has dramatically improved due to the adoption of deep, self-attentive architectures. However, these gains have come at the cost of hindering inference speed,…

Computation and Language · Computer Science 2020-10-30 Alexander Lin , Jeremy Wohlwend , Howard Chen , Tao Lei

In recent years, knowledge distillation (KD) has been widely used to derive efficient models. Through imitating a large teacher model, a lightweight student model can achieve comparable performance with more efficiency. However, most…

Computer Vision and Pattern Recognition · Computer Science 2023-03-23 Qizhen Lan , Qing Tian

We consider the problem of imitation learning from a finite set of expert trajectories, without access to reinforcement signals. The classical approach of extracting the expert's reward function via inverse reinforcement learning, followed…

Machine Learning · Computer Science 2019-06-10 Ruohan Wang , Carlo Ciliberto , Pierluigi Amadori , Yiannis Demiris

We introduce Proximal Policy Distillation (PPD), a novel policy distillation method that integrates student-driven distillation and Proximal Policy Optimization (PPO) to increase sample efficiency and to leverage the additional rewards that…

Machine Learning · Computer Science 2026-05-11 Giacomo Spigler

The privacy of machine learning models has become a significant concern in many emerging Machine-Learning-as-a-Service applications, where prediction services based on well-trained models are offered to users via pay-per-query. The lack of…

Machine Learning · Computer Science 2022-06-24 Xun Xian , Mingyi Hong , Jie Ding

Leveraging knowledge from multiple tasks through introducing a small number of task specific parameters into each transformer layer, also known as adapters, receives much attention recently. However, adding an extra fusion layer to…

Machine Learning · Computer Science 2023-12-29 Junjie Wang , Yicheng Chen , Wangshu Zhang , Sen Hu , Teng Xu , Jing Zheng

Despite that current reading comprehension systems have achieved significant advancements, their promising performances are often obtained at the cost of making an ensemble of numerous models. Besides, existing approaches are also…

Computation and Language · Computer Science 2018-09-18 Minghao Hu , Yuxing Peng , Furu Wei , Zhen Huang , Dongsheng Li , Nan Yang , Ming Zhou

Knowledge distillation is an effective method for training lightweight vision models. However, acquiring teacher supervision for training samples is often costly, especially from large-scale models like vision transformers (ViTs). In this…

Machine Learning · Computer Science 2024-09-30 Seungwoo Son , Jegwang Ryu , Namhoon Lee , Jaeho Lee

Autoregressive video generators are attractive for streaming, long-horizon, and interactive applications, but distilling strong black-box teachers into causal students remains difficult. The student must learn under its own rollout…

Computer Vision and Pattern Recognition · Computer Science 2026-05-26 Yang Luo , Shengju Qian , Xiaohang Tang , Zirui Zhu , Yong Liu , Xin Wang , Yang You

The reward model (RM), as the core component of reinforcement learning from human feedback (RLHF) for large language models (LLMs), responsible for providing reward signals to generated responses. However, the mainstream discriminative…

Computation and Language · Computer Science 2026-01-14 Jianxiang Zang