8 citations · 11 across the 10 of their papers we have counts for
11 papers
Token-Level Off-Policy Learning for Faithful Generation Under Distribution Shift
Zitong Huang, Gustavo Lucas Carvalho, Deqing Fu +1
We propose Token-Level Off-Policy Labeling (TOPL), an off-policy training paradigm that reframes post-training as a token-level correctness prediction task. Our key intuition is th…
CGL: Advancing Continual GUI Learning via Reinforcement Fine-Tuning
Zhenquan Yao, Zitong Huang, Yihan Zeng +5
Graphical User Interface (GUI) Agents, benefiting from recent advances in multimodal large language models (MLLM), have achieved significant development. However, due to the freque…
Mind the Generative Details: Direct Localized Detail Preference Optimization for Video Diffusion Models
Zitong Huang, Kaidong Zhang, Yukang Ding +4
Aligning text-to-video diffusion models with human preferences is crucial for generating high-quality videos. Existing Direct Preference Otimization (DPO) methods rely on multi-sam…
MIRAGE: Assessing Hallucination in Multimodal Reasoning Chains of MLLM
Bowen Dong, Minheng Ni, Zitong Huang +3
Multimodal hallucination in multimodal large language models (MLLMs) restricts the correctness of MLLMs. However, multimodal hallucinations are multi-sourced and arise from diverse…
Segmenting Objectiveness and Task-awareness Unknown Region for Autonomous Driving
Mi Zheng, Guanglei Yang, Zitong Huang +3
With the emergence of transformer-based architectures and large language models (LLMs), the accuracy of road scene perception has substantially advanced. Nonetheless, current road…
MR-GDINO: Efficient Open-World Continual Object Detection
Bowen Dong, Zitong Huang, Guanglei Yang +2
Open-world (OW) recognition and detection models show strong zero- and few-shot adaptation abilities, inspiring their use as initializations in continual learning methods to improv…
Class Balance Matters to Active Class-Incremental Learning
Zitong Huang, Ze Chen, Yuanze Li +6
Few-Shot Class-Incremental Learning has shown remarkable efficacy in efficient learning new concepts with limited annotations. Nevertheless, the heuristic few-shot annotations may…
CPT: Consistent Proxy Tuning for Black-box Optimization
Yuanyang He, Zitong Huang, Xinxing Xu +5
Black-box tuning has attracted recent attention due to that the structure or inner parameters of advanced proprietary models are not accessible. Proxy-tuning provides a test-time o…
LayerMatch: Do Pseudo-labels Benefit All Layers?
Chaoqi Liang, Guanglei Yang, Lifeng Qiao +4
Deep neural networks have achieved remarkable performance across various tasks when supplied with large-scale labeled data. However, the collection of labeled data can be time-cons…
IMWA: Iterative Model Weight Averaging Benefits Class-Imbalanced Learning Tasks
Zitong Huang, Ze Chen, Bowen Dong +3
Model Weight Averaging (MWA) is a technique that seeks to enhance model's performance by averaging the weights of multiple trained models. This paper first empirically finds that 1…
Learning Prompt with Distribution-Based Feature Replay for Few-Shot Class-Incremental Learning
Zitong Huang, Ze Chen, Zhixing Chen +6
Few-shot Class-Incremental Learning (FSCIL) aims to continuously learn new classes based on very limited training data without forgetting the old ones encountered. Existing studies…