Xueting Han

Haozhen Zhang, Xueting Han, Xi Xiao et al.

Temporal Graph Learning, which aims to model the time-evolving nature of graphs, has gained increasing attention and achieved remarkable performance recently. However, in reality, graph structures are often incomplete and noisy, which hinders temporal graph networks (TGNs) from learning informative representations. Graph contrastive learning uses data augmentation to generate plausible variations of existing data and learn robust representations. However, rule-based augmentation approaches may be suboptimal as they lack learnability and fail to leverage rich information from downstream tasks. To address these issues, we propose a Time-aware Graph Structure Learning (TGSL) approach via sequence prediction on temporal graphs, which learns better graph structures for downstream tasks through adding potential temporal edges. In particular, it predicts time-aware context embedding based on previously observed interactions and uses the Gumble-Top-K to select the closest candidate edges to this context embedding. Additionally, several candidate sampling strategies are proposed to ensure both efficiency and diversity. Furthermore, we jointly learn the graph structure and TGNs in an end-to-end manner and perform inference on the refined graph. Extensive experiments on temporal link prediction benchmarks demonstrate that TGSL yields significant gains for the popular TGNs such as TGAT and GraphMixer, and it outperforms other contrastive learning methods on temporal graphs. We release the code at https://github.com/ViktorAxelsen/TGSL.

Shengrui Li, Xueting Han, Jing Bai

Fine-tuning pre-trained models has recently yielded remarkable performance gains in graph neural networks (GNNs). In addition to pre-training techniques, inspired by the latest work in the natural language fields, more recent work has shifted towards applying effective fine-tuning approaches, such as parameter-efficient fine-tuning (PEFT). However, given the substantial differences between GNNs and transformer-based models, applying such approaches directly to GNNs proved to be less effective. In this paper, we present a comprehensive comparison of PEFT techniques for GNNs and propose a novel PEFT method specifically designed for GNNs, called AdapterGNN. AdapterGNN preserves the knowledge of the large pre-trained model and leverages highly expressive adapters for GNNs, which can adapt to downstream tasks effectively with only a few parameters, while also improving the model's generalization ability. Extensive experiments show that AdapterGNN achieves higher performance than other PEFT methods and is the only one consistently surpassing full fine-tuning (outperforming it by 1.6% and 5.7% in the chemistry and biology domains respectively, with only 5% and 4% of its parameters tuned) with lower generalization gaps. Moreover, we empirically show that a larger GNN model can have a worse generalization ability, which differs from the trend observed in large transformer-based models. Building upon this, we provide a theoretical justification for PEFT can improve generalization of GNNs by applying generalization bounds. Our code is available at https://github.com/Lucius-lsr/AdapterGNN.

Dongshuo Yin, Xueting Han, Bin Li et al.

Pre-training & fine-tuning is a prevalent paradigm in computer vision (CV). Recently, parameter-efficient transfer learning (PETL) methods have shown promising performance in adapting to downstream tasks with only a few trainable parameters. Despite their success, the existing PETL methods in CV can be computationally expensive and require large amounts of memory and time cost during training, which limits low-resource users from conducting research and applications on large models. In this work, we propose Parameter, Memory, and Time Efficient Visual Adapter ($\mathrm{E^3VA}$) tuning to address this issue. We provide a gradient backpropagation highway for low-rank adapters which eliminates the need for expensive backpropagation through the frozen pre-trained model, resulting in substantial savings of training memory and training time. Furthermore, we optimise the $\mathrm{E^3VA}$ structure for CV tasks to promote model performance. Extensive experiments on COCO, ADE20K, and Pascal VOC benchmarks show that $\mathrm{E^3VA}$ can save up to 62.2% training memory and 26.2% training time on average, while achieving comparable performance to full fine-tuning and better performance than most PETL methods. Note that we can even train the Swin-Large-based Cascade Mask RCNN on GTX 1080Ti GPUs with less than 1.5% trainable parameters.

Tianlong Wang, Junzhe Chen, Xueting Han et al.

Post-training, particularly reinforcement learning (RL) using self-play-generated data, has become a new learning paradigm for large language models (LLMs). However, scaling RL to develop a general reasoner remains a research challenge, as existing methods focus on task-specific reasoning without adequately addressing generalization across a broader range of tasks. Moreover, unlike traditional RL with limited action space, LLMs operate in an infinite space, making it crucial to search for valuable and diverse strategies to solve problems effectively. To address this, we propose searching within the action space on high-level abstract plans to enhance model generalization and introduce Critical Plan Step Learning (CPL), comprising: 1) searching on plan, using Monte Carlo Tree Search (MCTS) to explore diverse plan steps in multi-step reasoning tasks, and 2) learning critical plan steps through Step-level Advantage Preference Optimization (Step-APO), which integrates advantage estimates for step preference obtained via MCTS into Direct Preference Optimization (DPO). This combination helps the model effectively learn critical plan steps, enhancing both reasoning capabilities and generalization. Experimental results demonstrate that our method, trained exclusively on GSM8K and MATH, not only significantly improves performance on GSM8K (+10.5%) and MATH (+6.5%), but also enhances out-of-domain reasoning benchmarks, such as HumanEval (+12.2%), GPQA (+8.6%), ARC-C (+4.0%), MMLU-STEM (+2.2%), and BBH (+1.8%).

Shengrui Li, Junzhe Chen, Xueting Han et al.

The considerable size of Large Language Models (LLMs) presents notable deployment challenges, particularly on resource-constrained hardware. Structured pruning, offers an effective means to compress LLMs, thereby reducing storage costs and enhancing inference speed for more efficient utilization. In this work, we study data-efficient and resource-efficient structure pruning methods to obtain smaller yet still powerful models. Knowledge Distillation is well-suited for pruning, as the intact model can serve as an excellent teacher for pruned students. However, it becomes challenging in the context of LLMs due to memory constraints. To address this, we propose an efficient progressive Numerous-teacher pruning method (NutePrune). NutePrune mitigates excessive memory costs by loading only one intact model and integrating it with various masks and LoRA modules, enabling it to seamlessly switch between teacher and student roles. This approach allows us to leverage numerous teachers with varying capacities to progressively guide the pruned model, enhancing overall performance. Extensive experiments across various tasks demonstrate the effectiveness of NutePrune. In LLaMA-7B zero-shot experiments, NutePrune retains 97.17% of the performance of the original model at 20% sparsity and 95.07% at 25% sparsity. Our code is available at https://github.com/Lucius-lsr/NutePrune.

Liang Chen, Xueting Han, Li Shen et al.

Harmful fine-tuning (HFT), performed directly on open-source LLMs or through Fine-tuning-as-a-Service, breaks safety alignment and poses significant threats. Existing methods aim to mitigate HFT risks by learning robust representation on alignment data or making harmful data unlearnable, but they treat each data sample equally, leaving data vulnerability patterns understudied. In this work, we reveal that certain subsets of alignment data are consistently more prone to forgetting during HFT across different fine-tuning tasks. Inspired by these findings, we propose Vulnerability-Aware Alignment (VAA), which estimates data vulnerability, partitions data into "vulnerable" and "invulnerable" groups, and encourages balanced learning using a group distributionally robust optimization (Group DRO) framework. Specifically, VAA learns an adversarial sampler that samples examples from the currently underperforming group and then applies group-dependent adversarial perturbations to the data during training, aiming to encourage a balanced learning process across groups. Experiments across four fine-tuning tasks demonstrate that VAA significantly reduces harmful scores while preserving downstream task performance, outperforming state-of-the-art baselines.

Ke Ma, Jiaqi Tang, Bin Guo et al.

Proactive streaming video understanding requires Video-LLMs to decide when to respond as a video unfolds, a task where existing methods often fall short due to their implicit, query-agnostic modeling of visual evidence. We introduce Response-G1, a novel framework that establishes explicit, structured alignment between the accumulated video evidence and the query's expected response conditions via scene graphs. The framework operates in three fine-tuning-free stages: (1) online query-guided scene graph generation from streaming clips; (2) memory-based retrieval of the most semantically relevant historical scene graphs; and (3) retrieval-augmented trigger prompting for per-frame "silence/response" decisions.By grounding both evidence and conditions in a shared graph representation, Response-G1 achieves more interpretable and accurate response timing decisions. Experimental results on established benchmarks demonstrate the superiority of our method in both proactive and reactive tasks, validating the advantage of explicit scene graph modeling and retrieval in streaming video understanding.

Liang Chen, Xueting Han, Li Shen et al.

Reinforcement learning (RL) has proven effective in incentivizing the reasoning abilities of large language models (LLMs), but suffers from severe efficiency challenges due to its trial-and-error nature. While the common practice employs supervised fine-tuning (SFT) as a warm-up stage for RL, this decoupled two-stage approach suffers from catastrophic forgetting: second-stage RL gradually loses SFT-acquired behaviors and inefficiently explores new patterns. This study introduces a novel method for learning reasoning models that employs bilevel optimization to facilitate better cooperation between these training paradigms. By conditioning the SFT objective on the optimal RL policy, our approach enables SFT to meta-learn how to guide RL's optimization process. During training, the lower level performs RL updates while simultaneously receiving SFT supervision, and the upper level explicitly maximizes the cooperative gain-the performance advantage of joint SFT-RL training over RL alone. Empirical evaluations on five reasoning benchmarks demonstrate that our method consistently outperforms baselines and achieves a better balance between effectiveness and efficiency.

Yiyang Jin, Kunzhao Xu, Hang Li et al.

Reinforcement learning with verifiable rewards (RLVR) has advanced the reasoning capabilities of large language models. However, existing methods rely solely on outcome rewards, without explicitly optimizing verification or leveraging reliable signals from realistic environments, leading to unreliable self-verification and limited test-time scaling. To address this, we widen the verification-generation asymmetry by explicitly optimizing self-verification, making it a reliable driver of deeper test-time scaling. We introduce ReVeal, a multi-turn reinforcement learning framework that evolves code generation through self-verification and tool-based evaluation. ReVeal structures long-horizon reasoning as iterative generation-verification turns and incorporates TAPO for turn-level credit assignment, fostering the co-evolution of code and test generation. At inference, this strengthened self-verification enables the model to use self-constructed tests and tool feedback to continuously evolve code for 20+ turns on LiveCodeBench despite training on only three. It also significantly improves Pass@k, indicating stronger exploration that expands the reasoning boundaries of the base model. These findings highlight the promise of ReVeal as a scalable paradigm for RL training and test-time scaling, paving the way for more robust and autonomous AI agents.

Weixi Xiang, Xueting Han, Xiujuan Chai et al.

Modeling biological sequences such as DNA, RNA, and proteins is crucial for understanding complex processes like gene regulation and protein synthesis. However, most current models either focus on a single type or treat multiple types of data separately, limiting their ability to capture cross-modal relationships. We propose that by learning the relationships between these modalities, the model can enhance its understanding of each type. To address this, we introduce BSM, a small but powerful mixed-modal biological sequence foundation model, trained on three types of data: RefSeq, Gene Related Sequences, and interleaved biological sequences from the web. These datasets capture the genetic flow, gene-protein relationships, and the natural co-occurrence of diverse biological data, respectively. By training on mixed-modal data, BSM significantly enhances learning efficiency and cross-modal representation, outperforming models trained solely on unimodal data. With only 110M parameters, BSM achieves performance comparable to much larger models across both single-modal and mixed-modal tasks, and uniquely demonstrates in-context learning capability for mixed-modal tasks, which is absent in existing models. Further scaling to 270M parameters demonstrates even greater performance gains, highlighting the potential of BSM as a significant advancement in multimodal biological sequence modeling.

Liang Chen, Xueting Han, Qizhou Wang et al.

Balancing exploration and exploitation remains a central challenge in reinforcement learning with verifiable rewards (RLVR) for large language models (LLMs). Current RLVR methods often overemphasize exploitation, leading to entropy collapse, diminished exploratory capacity, and ultimately limited performance gains. Although techniques that increase policy stochasticity can promote exploration, they frequently fail to escape dominant behavioral modes. This creates a self-reinforcing loop-repeatedly sampling and rewarding dominant modes-that further erodes exploration. We introduce Exploration-Enhanced Policy Optimization (EEPO), a framework that promotes exploration via two-stage rollouts with adaptive unlearning. In the first stage, the model generates half of the trajectories; it then undergoes a lightweight unlearning step to temporarily suppress these sampled responses, forcing the second stage to explore different regions of the output space. This sample-then-forget mechanism disrupts the self-reinforcing loop and promotes wider exploration during rollouts. Across five reasoning benchmarks, EEPO outperforms GRPO, achieving average relative gains of 24.3% on Qwen2.5-3B, 33.0% on Llama3.2-3B-Instruct, and 10.4% on Qwen3-8B-Base.

Xueting Han, Zhenhuan Huang, Bang An et al.

Graph neural networks (GNNs) is widely used to learn a powerful representation of graph-structured data. Recent work demonstrates that transferring knowledge from self-supervised tasks to downstream tasks could further improve graph representation. However, there is an inherent gap between self-supervised tasks and downstream tasks in terms of optimization objective and training data. Conventional pre-training methods may be not effective enough on knowledge transfer since they do not make any adaptation for downstream tasks. To solve such problems, we propose a new transfer learning paradigm on GNNs which could effectively leverage self-supervised tasks as auxiliary tasks to help the target task. Our methods would adaptively select and combine different auxiliary tasks with the target task in the fine-tuning stage. We design an adaptive auxiliary loss weighting model to learn the weights of auxiliary tasks by quantifying the consistency between auxiliary tasks and the target task. In addition, we learn the weighting model through meta-learning. Our methods can be applied to various transfer learning approaches, it performs well not only in multi-task learning but also in pre-training and fine-tuning. Comprehensive experiments on multiple downstream tasks demonstrate that the proposed methods can effectively combine auxiliary tasks with the target task and significantly improve the performance compared to state-of-the-art methods.