5 accepted papers
Existing alignment methods directly use the reward model learned from user preference data to optimize an LLM policy, subject to KL regularization with respect to the base policy. This practice is suboptimal for maximizing user's utility because the KL regularization may cause the LLM to inherit the…
Generating high-quality synonymous perturbations is a core challenge for textual adversarial tasks. However, candidates generated from the masked language model often contain many words that are antonyms or irrelevant to the original words, which limit the perturbation space and affect the attack’s…
Recently, many excellent weakly supervised semantic segmentation (WSSS) works are proposed based on class activation mapping (CAM). However, there are few works that consider the characteristics of medical images. In this paper, we find that there are mainly two challenges of medical images in WSSS:…
Despite great effectiveness of very deep and wide Convolutional Neural Networks (CNNs) in various computer vision tasks, the significant cost in terms of storage requirement of such networks impedes the deployment on computationally limited devices. In this paper, we propose new Modulated Convolutio…
Trajectory classification has been extensively investigated in recent years, however, the problems about automatically modeling unlabeled and incomplete trajectories are far from being solved. In this paper, we propose a Cluster-based Dictionary Learning (CDL) approach that firstly constructs an ini…