13 accepted papers
Group-based reinforcement learning (RL) methods have achieved remarkable success in improving the performance of large language models (LLMs) and have been rapidly extended to agentic tasks. However, their credit assignment relies heavily on coarse-grained trajectory-level attribution according to f…
Group-based reinforcement learning (RL), such as GRPO, has advanced the capabilities of large language models on long-horizon agentic tasks. To enable more fine-grained policy updates, recent research has increasingly shifted toward stepwise group-based policy optimization, which treats each step in…
Reinforcement learning (RL) has equipped LLM agents with a strong ability to solve complex tasks. However, existing RL methods normally use a single policy network, causing simplicity bias where simple tasks occupy most parameters and dominate gradient updates, leaving insufficient capacity for comp…
Despite the strong multimodal performance, large vision-language models (LVLMs) are vulnerable during fine-tuning to backdoor attacks, where adversaries insert trigger-embedded samples into the training data to implant behaviors that can be maliciously activated at test time. Existing defenses typic
We present a comprehensive empirical study on how backdoor attacks affect CLIP by analyzing the representations of backdoor images. Specifically, based on the methodology of representation decomposing, image representations can be decomposed into a sum of representations across individual image patc…
Multimodal contrastive learning models (e.g., CLIP) can learn high-quality representations from large-scale image-text datasets, while they exhibit significant vulnerabilities to backdoor attacks, raising serious safety concerns. In this paper, we reveal that CLIP's vulnerabilities primarily stem fr…
Reinforcement Learning (RL), trained via trial and error in simulators, has been proven to be an effective approach for addressing task assignment problems in spatial crowdsourcing. However, a performance gap still exists when transferring the simulator-trained RL Models (RLMs) to real-world setting
Backdoor attacks have posed a serious threat in machine learning models, wherein adversaries can poison training samples with maliciously crafted triggers to compromise the victim model. Advanced backdoor attack methods have focused on selectively poisoning more vulnerable training samples, achievin…
Model merging aims to achieve multitask performance by merging multiple expert models without the need to access the raw training data. Recent research identified the \textit{representation bias} of model merging, characterized by a discrepancy in the representation distribution between the merged a…
While multimodal contrastive learning methods (e.g., CLIP) can achieve impressive zero-shot classification performance, recent research has revealed that these methods are vulnerable to backdoor attacks. To defend against backdoor attacks on CLIP, existing defense methods focus on either the pre-tra…
Partial-label learning (PLL) allows each training example to be equipped with a set of candidate labels. Existing deep PLL research focuses on a \emph{learning-centric} perspective to design various training strategies for label disambiguation i.e., identifying the concealed true label from the cand…
In contrast to the standard learning paradigm where all classes can be observed in training data, learning with augmented classes (LAC) tackles the problem where augmented classes unobserved in the training data may emerge in the test phase. Previous research showed that given unlabeled data, an unb…
In partial-label learning (PLL), each training example has a set of candidate labels, among which only one is the true label. Most existing PLL studies focus on the instance-independent (II) case, where the generation of candidate labels is only dependent on the true label. However, this II-PLL para…