Publications (113)
Ai2-Kit: Streamlining AI-Accelerated Ab Initio Workflows for Complex Chemical Systems
Sheng Bi, Wei-Hong Xu, Yong-Bin Zhuang +47
The paper introduces ai2-kit, a software toolkit that streamlines AI‑accelerated ab initio workflows for complex chemical systems by providing command‑line and Python interfaces fo…
Adaptive Spatial Modulation for Visible Light Communications with an Arbitrary Number of Transmitters
Jin-Yuan Wang, Hong Ge, Jian-Xia Zhu +3
As a power and bandwidth efficient modulation scheme, the optical spatial modulation (SM) technique has recently drawn increased attention in the field of visible light communicati…
Chain of Preference Optimization: Improving Chain-of-Thought Reasoning in LLMs
Xuan Zhang, Chao Du, Tianyu Pang +3
The recent development of chain-of-thought (CoT) decoding has enabled large language models (LLMs) to generate explicit logical reasoning paths for complex problem-solving. However…
Network In Network
Min Lin, Qiang Chen, Shuicheng Yan
We propose a novel deep network structure called "Network In Network" (NIN) to enhance model discriminability for local patches within the receptive field. The conventional convolu…
Purifying Large Language Models by Ensembling a Small Language Model
Tianlin Li, Qian Liu, Tianyu Pang +4
The emerging success of large language models (LLMs) heavily relies on collecting abundant training data from external (untrusted) sources. Despite substantial efforts devoted to d…
Purine: A bi-graph based deep learning framework
Min Lin, Shuo Li, Xuan Luo +1
In this paper, we introduce a novel deep learning framework, termed Purine. In Purine, a deep network is expressed as a bipartite graph (bi-graph), which is composed of interconnec…
Agent Smith: A Single Image Can Jailbreak One Million Multimodal LLM Agents Exponentially Fast
Xiangming Gu, Xiaosen Zheng, Tianyu Pang +5
A multimodal large language model (MLLM) agent can receive instructions, capture images, retrieve histories from memory, and decide which tools to use. Nonetheless, red-teaming eff…
Sailor2: Sailing in South-East Asia with Inclusive Multilingual LLMs
Longxu Dou, Qian Liu, Fan Zhou +38
Sailor2 is a family of cutting-edge multilingual language models for South-East Asian (SEA) languages, available in 1B, 8B, and 20B sizes to suit diverse applications. Building on…
Beyond Memorization: The Challenge of Random Memory Access in Language Models
Tongyao Zhu, Qian Liu, Liang Pang +3
Recent developments in Language Models (LMs) have shown their effectiveness in NLP tasks, particularly in knowledge-intensive tasks. However, the mechanisms underlying knowledge st…
PhyBlock: A Progressive Benchmark for Physical Understanding and Planning via 3D Block Assembly
Liang Ma, Jiajun Wen, Min Lin +12
While vision-language models (VLMs) have demonstrated promising capabilities in reasoning and planning for embodied agents, their ability to comprehend physical phenomena, particul…
AI4X Roadmap: Artificial Intelligence for the advancement of scientific pursuit and its future directions
Stephen G. Dale, Nikita Kazeev, Alastair J. A. Price +65
Artificial intelligence and machine learning are reshaping how we approach scientific discovery, not by replacing established methods but by extending what researchers can probe, p…
Defeating the Training-Inference Mismatch via FP16
Penghui Qi, Zichen Liu, Xiangxin Zhou +4
Reinforcement learning (RL) fine-tuning of large language models (LLMs) often suffers from instability due to the numerical mismatch between the training and inference policies. Wh…
Scaling up Masked Diffusion Models on Text
Shen Nie, Fengqi Zhu, Chao Du +5
Masked diffusion models (MDMs) have shown promise in language modeling, yet their scalability and effectiveness in core language tasks, such as text generation and language underst…
Continual Learning from the Perspective of Compression
Xu He, Min Lin
Connectionist models such as neural networks suffer from catastrophic forgetting. In this work, we study this problem from the perspective of information theory and define forgetti…
Physical-layer Security for Indoor Visible Light Communications: Secrecy Capacity Analysis
Jin-Yuan Wang, Cheng Liu, Jun-Bo Wang +3
This paper investigates the physical-layer security for an indoor visible light communication (VLC) network consisting of a transmitter, a legitimate receiver and an eavesdropper.…
Robustness and Accuracy Could Be Reconcilable by (Proper) Definition
Tianyu Pang, Min Lin, Xiao Yang +2
The trade-off between robustness and accuracy has been widely studied in the adversarial literature. Although still controversial, the prevailing view is that this trade-off is inh…
Graph Diffusion Policy Optimization
Yijing Liu, Chao Du, Tianyu Pang +3
Recent research has made significant progress in optimizing diffusion models for downstream objectives, which is an important pursuit in fields such as graph generation for drug de…
Joint Beamforming and Computation Offloading for Multi-user Mobile-Edge Computing
Changfeng Ding, Jun-Bo Wang, Ming Cheng +3
Mobile edge computing (MEC) is considered as an efficient method to relieve the computation burden of mobile devices. In order to reduce the energy consumption and time delay of mo…
Meta-Unlearning on Diffusion Models: Preventing Relearning Unlearned Concepts
Hongcheng Gao, Tianyu Pang, Chao Du +3
With the rapid progress of diffusion-based content generation, significant efforts are being made to unlearn harmful or copyrighted concepts from pretrained diffusion models (DMs)…
Hovering UAV-Based FSO Communications: Channel Modelling, Performance Analysis, and Parameter Optimization
Jin-Yuan Wang, Yang Ma, Rong-Rong Lu +3
Relay-assisted free-space optical (FSO) communication systems are exploited as a means to mitigate the limiting effects of the turbulence induced atmospheric scintillation. However…
Simultaneous confidence bands for cumulative hazard via exchangeable bootstrap and box calibration
Min Lin, Grzegorz Rempala, Eben Kenah +1
Resampling-based simultaneous confidence bands for cumulative hazard functions often undercover in finite samples with right censoring. We study two aspects of the construction tha…
Zero Bubble Pipeline Parallelism
Penghui Qi, Xinyi Wan, Guangxing Huang +1
Pipeline parallelism is one of the key components for large-scale distributed training, yet its efficiency suffers from pipeline bubbles which were deemed inevitable. In this work,…
Optimizing Anytime Reasoning via Budget Relative Policy Optimization
Penghui Qi, Zichen Liu, Tianyu Pang +3
Scaling test-time compute is crucial for enhancing the reasoning capabilities of large language models (LLMs). Existing approaches typically employ reinforcement learning (RL) to m…
Gradient based sample selection for online continual learning
Rahaf Aljundi, Min Lin, Baptiste Goujaud +1
A continual learning agent learns online with a non-stationary and never-ending stream of data. The key to such learning process is to overcome the catastrophic forgetting of previ…
Intriguing Properties of Data Attribution on Diffusion Models
Xiaosen Zheng, Tianyu Pang, Chao Du +2
Data attribution seeks to trace model outputs back to training data. With the recent development of diffusion models, data attribution has become a desired module to properly assig…
On Memorization in Diffusion Models
Xiangming Gu, Chao Du, Tianyu Pang +3
Due to their capacity to generate novel and high-quality samples, diffusion models have attracted significant research interest in recent years. Notably, the typical training objec…
Rethinking the Trust Region in LLM Reinforcement Learning
Penghui Qi, Xiangxin Zhou, Zichen Liu +4
Reinforcement learning (RL) has become a cornerstone for fine-tuning Large Language Models (LLMs), with Proximal Policy Optimization (PPO) serving as the de facto standard algorith…
Benchmarking Large Multimodal Models against Common Corruptions
Jiawei Zhang, Tianyu Pang, Chao Du +3
This technical report aims to fill a deficiency in the assessment of large multimodal models (LMMs) by specifically examining the self-consistency of their outputs when subjected t…
Locality Sensitive Sparse Encoding for Learning World Models Online
Zichen Liu, Chao Du, Wee Sun Lee +1
Acquiring an accurate world model online for model-based reinforcement learning (MBRL) is challenging due to data nonstationarity, which typically causes catastrophic forgetting fo…
FlowMamba: Learning Point Cloud Scene Flow with Global Motion Propagation
Min Lin, Gangwei Xu, Yun Wang +2
Scene flow methods based on deep learning have achieved impressive performance. However, current top-performing methods still struggle with ill-posed regions, such as extensive fla…
Photonic quasicrystal of spin angular momentum
Min Lin, Xinxin Gou, Zhenwei Xie +3
Quasicrystals,characterized by long-range order without translational symmetry,have catalyzed transformative advances in various fields,including optics in terms of field quasicrys…
Stochastic Taylor Derivative Estimator: Efficient amortization for arbitrary differential operators
Zekun Shi, Zheyuan Hu, Min Lin +1
Optimizing neural networks with loss that contain high-dimensional and high-order differential operators is expensive to evaluate with back-propagation due to …
Pipeline Parallelism with Controllable Memory
Penghui Qi, Xinyi Wan, Nyamdavaa Amar +1
Pipeline parallelism has been widely explored, but most existing schedules lack a systematic methodology. In this paper, we propose a framework to decompose pipeline schedules as r…
EnvPool: A Highly Parallel Reinforcement Learning Environment Execution Engine
Jiayi Weng, Min Lin, Shengyi Huang +9
There has been significant progress in developing reinforcement learning (RL) training systems. Past works such as IMPALA, Apex, Seed RL, Sample Factory, and others, aim to improve…
Bandit Inspired Beam Searching Scheme for mmWave High-Speed Train Communications
Jun-Bo Wang, Ming Cheng, Jin-Yuan Wang +4
High-speed trains (HSTs) are being widely deployed around the world. To meet the high-rate data transmission requirements on HSTs, millimeter wave (mmWave) HST communications have…
D4FT: A Deep Learning Approach to Kohn-Sham Density Functional Theory
Tianbo Li, Min Lin, Zheyuan Hu +6
Kohn-Sham Density Functional Theory (KS-DFT) has been traditionally solved by the Self-Consistent Field (SCF) method. Behind the SCF loop is the physics intuition of solving a syst…
Outage Constrained Robust Secure Beamforming in Cognitive Satellite-Aerial Networks
Bai Zhao, Min Lin, Ming Cheng +2
This paper proposes a robust beamforming scheme to enhance the physical layer security (PLS) of multicast transmission in a cognitive satellite and aerial network (CSAN) operating…
Understanding R1-Zero-Like Training: A Critical Perspective
Zichen Liu, Changyu Chen, Wenjun Li +5
DeepSeek-R1-Zero has shown that reinforcement learning (RL) at scale can directly enhance the reasoning capabilities of LLMs without supervised fine-tuning. In this work, we critic…
Instant3D: Instant Text-to-3D Generation
Ming Li, Pan Zhou, Jia-Wei Liu +4
Text-to-3D generation has attracted much attention from the computer vision community. Existing methods mainly optimize a neural field from scratch for each text prompt, relying on…
Improving Your Model Ranking on Chatbot Arena by Vote Rigging
Rui Min, Tianyu Pang, Chao Du +3
Chatbot Arena is a popular platform for evaluating LLMs by pairwise battles, where users vote for their preferred response from two randomly sampled anonymous models. While Chatbot…
ZeroStereo: Zero-shot Stereo Matching from Single Images
Xianqi Wang, Hao Yang, Gangwei Xu +6
State-of-the-art supervised stereo matching methods have achieved remarkable performance on various benchmarks. However, their generalization to real-world scenarios remains challe…
On the Secrecy Rate of Spatial Modulation Based Indoor Visible Light Communications
Jin-Yuan Wang, Hong Ge, Min Lin +3
In this paper, we investigate the physical-layer security for a spatial modulation (SM) based indoor visible light communication (VLC) system, which includes multiple transmitters,…
Sample-Efficient Alignment for LLMs
Zichen Liu, Changyu Chen, Chao Du +2
We study methods for efficiently aligning large language models (LLMs) with human preferences given budgeted online feedback. We first formulate the LLM alignment problem in the fr…
RegMix: Data Mixture as Regression for Language Model Pre-training
Qian Liu, Xiaosen Zheng, Niklas Muennighoff +5
The data mixture for large language model pre-training significantly impacts performance, yet how to determine an effective mixture remains unclear. We propose RegMix to automatica…
Sailor: Open Language Models for South-East Asia
Longxu Dou, Qian Liu, Guangtao Zeng +4
We present Sailor, a family of open language models ranging from 0.5B to 7B parameters, tailored for South-East Asian (SEA) languages. These models are continually pre-trained from…
A Recipe for Watermarking Diffusion Models
Yunqing Zhao, Tianyu Pang, Chao Du +3
Diffusion models (DMs) have demonstrated advantageous potential on generative tasks. Widespread interest exists in incorporating DMs into downstream applications, such as producing…
Cheating Automatic LLM Benchmarks: Null Models Achieve High Win Rates
Xiaosen Zheng, Tianyu Pang, Chao Du +3
Automatic LLM benchmarks, such as AlpacaEval 2.0, Arena-Hard-Auto, and MT-Bench, have become popular for evaluating language models due to their cost-effectiveness and scalability…
Exploring Incompatible Knowledge Transfer in Few-shot Image Generation
Yunqing Zhao, Chao Du, Milad Abdollahzadeh +4
Few-shot image generation (FSIG) learns to generate diverse and high-fidelity images from a target domain using a few (e.g., 10) reference samples. Existing FSIG methods select, pr…
FlowReasoner: Reinforcing Query-Level Meta-Agents
Hongcheng Gao, Yue Liu, Yufei He +6
This paper proposes a query-level meta-agent named FlowReasoner to automate the design of query-level multi-agent systems, i.e., one system per user query. Our core idea is to ince…
Nonparametric Data Attribution for Diffusion Models
Yutian Zhao, Chao Du, Xiaosen Zheng +2
Data attribution for generative models seeks to quantify the influence of individual training examples on model outputs. Existing methods for diffusion models typically require acc…
Reinforcing General Reasoning without Verifiers
Xiangxin Zhou, Zichen Liu, Anya Sims +6
The recent paradigm shift towards training large language models (LLMs) using DeepSeek-R1-Zero-style reinforcement learning (RL) on verifiable rewards has led to impressive advance…
Automatic Functional Differentiation in JAX
Min Lin
We extend JAX with the capability to automatically differentiate higher-order functions (functionals and operators). By representing functions as a generalization of arrays, we sea…
Online Fast Adaptation and Knowledge Accumulation: a New Approach to Continual Learning
Massimo Caccia, Pau Rodriguez, Oleksiy Ostapenko +8
Continual learning studies agents that learn from streams of tasks without forgetting previous ones while adapting to new ones. Two recent continual-learning scenarios have opened…
Revisiting Parameter Server in LLM Post-Training
Xinyi Wan, Penghui Qi, Guangxing Huang +3
Modern data parallel (DP) training favors collective communication over parameter servers (PS) for its simplicity and efficiency under balanced workloads. However, the balanced wor…
Scaling Laws with Vocabulary: Larger Models Deserve Larger Vocabularies
Chaofan Tao, Qian Liu, Longxu Dou +5
Research on scaling large language models (LLMs) has primarily focused on model parameters and training data size, overlooking the role of vocabulary size. We investigate how vocab…
Causal Attention for Interpretable and Generalizable Graph Classification
Yongduo Sui, Xiang Wang, Jiancan Wu +3
In graph classification, attention and pooling-based graph neural networks (GNNs) prevail to extract the critical features from the input graph and support the prediction. They mos…
PipeOffload: Improving Scalability of Pipeline Parallelism with Memory Optimization
Xinyi Wan, Penghui Qi, Guangxing Huang +2
Pipeline parallelism (PP) is widely used for training large language models (LLMs), yet its scalability is often constrained by high activation memory consumption as the number of…
NU-MCC: Multiview Compressive Coding with Neighborhood Decoder and Repulsive UDF
Stefan Lionar, Xiangyu Xu, Min Lin +1
Remarkable progress has been made in 3D reconstruction from single-view RGB-D inputs. MCC is the current state-of-the-art method in this field, which achieves unprecedented success…
Improved Techniques for Optimization-Based Jailbreaking on Large Language Models
Xiaojun Jia, Tianyu Pang, Chao Du +5
Large language models (LLMs) are being rapidly developed, and a key component of their widespread deployment is their safety-related alignment. Many red-teaming efforts aim to jail…
Online Continual Learning with Maximally Interfered Retrieval
Rahaf Aljundi, Lucas Caccia, Eugene Belilovsky +4
Continual learning, the setting where a learning agent is faced with a never ending stream of data, continues to be a great challenge for modern machine learning systems. In partic…
On Evaluating Adversarial Robustness of Large Vision-Language Models
Yunqing Zhao, Tianyu Pang, Chao Du +4
Large vision-language models (VLMs) such as GPT-4 have achieved unprecedented performance in response generation, especially with visual inputs, enabling more creative and adaptabl…
How Should Pre-Trained Language Models Be Fine-Tuned Towards Adversarial Robustness?
Xinhsuai Dong, Luu Anh Tuan, Min Lin +2
The fine-tuning of pre-trained language models has a great success in many NLP fields. Yet, it is strikingly vulnerable to adversarial examples, e.g., word substitution attacks usi…
A Closer Look at Machine Unlearning for Large Language Models
Xiaojian Yuan, Tianyu Pang, Chao Du +3
Large language models (LLMs) may memorize sensitive or copyrighted content, raising privacy and legal concerns. Due to the high cost of retraining from scratch, researchers attempt…
From Zero to Hero: Examining the Power of Symbolic Tasks in Instruction Tuning
Qian Liu, Fan Zhou, Zhengbao Jiang +2
Fine-tuning language models on tasks with instructions has demonstrated potential in facilitating zero-shot generalization to unseen tasks. In this paper, we introduce a straightfo…
Structured Preference Optimization for Vision-Language Long-Horizon Task Planning
Xiwen Liang, Min Lin, Weiqi Ruan +6
Existing methods for vision-language task planning excel in short-horizon tasks but often fall short in complex, long-horizon planning within dynamic environments. These challenges…
Conditional Computation for Continual Learning
Min Lin, Jie Fu, Yoshua Bengio
Catastrophic forgetting of connectionist neural networks is caused by the global sharing of parameters among all training examples. In this study, we analyze parameter sharing unde…
When Attention Sink Emerges in Language Models: An Empirical View
Xiangming Gu, Tianyu Pang, Chao Du +5
Language Models (LMs) assign significant attention to the first token, even if it is not semantically important, which is known as attention sink. This phenomenon has been widely a…
Universal Antisymmetry in Fermionic Neural Networks
Tianyu Pang, Shuicheng Yan, Min Lin
Fermionic neural network (FermiNet) is a recently proposed wavefunction Ansatz, which is used in variational Monte Carlo (VMC) methods to solve the many-electron Schrödinger equat…
Finetuning Text-to-Image Diffusion Models for Fairness
Xudong Shen, Chao Du, Tianyu Pang +3
The rapid adoption of text-to-image diffusion models in society underscores an urgent need to address their biases. Without interventions, these biases could propagate a skewed wor…
Cleanba: A Reproducible and Efficient Distributed Reinforcement Learning Platform
Shengyi Huang, Jiayi Weng, Rujikorn Charakorn +3
Distributed Deep Reinforcement Learning (DRL) aims to leverage more computational resources to train autonomous agents with less training time. Despite recent progress in the field…
PromptStereo: Zero-Shot Stereo Matching via Structure and Motion Prompts
Xianqi Wang, Hao Yang, Hangtian Wang +4
Modern stereo matching methods have leveraged monocular depth foundation models to achieve superior zero-shot generalization performance. However, most existing methods primarily f…
BAFFLE: A Baseline of Backpropagation-Free Federated Learning
Haozhe Feng, Tianyu Pang, Chao Du +3
Federated learning (FL) is a general principle for decentralized clients to train a server model collectively without sharing local data. FL is a promising framework with practical…
A Machine Learning Framework for Resource Allocation Assisted by Cloud Computing
Jun-Bo Wang, Junyuan Wang, Yongpeng Wu +4
Conventionally, the resource allocation is formulated as an optimization problem and solved online with instantaneous scenario information. Since most resource allocation problems…
Nonparametric Generative Modeling with Conditional Sliced-Wasserstein Flows
Chao Du, Tianbo Li, Tianyu Pang +2
Sliced-Wasserstein Flow (SWF) is a promising approach to nonparametric generative modeling but has not been widely adopted due to its suboptimal generative quality and lack of cond…
Diagonalization without Diagonalization: A Direct Optimization Approach for Solid-State Density Functional Theory
Tianbo Li, Min Lin, Stephen Dale +4
We present a novel approach to address the challenges of variable occupation numbers in direct optimization of density functional theory (DFT). By parameterizing both the eigenfunc…
Bag of Tricks for Training Data Extraction from Language Models
Weichen Yu, Tianyu Pang, Qian Liu +5
With the advance of language models, privacy protection is receiving more attention. Training data extraction is therefore of great importance, as it can serve as a potential tool…
LSTM-RPA: A Simple but Effective Long Sequence Prediction Algorithm for Music Popularity Prediction
Kun Li, Meng Li, Yanling Li +1
The big data about music history contains information about time and users' behavior. Researchers could predict the trend of popular songs accurately by analyzing this data. The tr…
Softmax GAN
Min Lin
Softmax GAN is a novel variant of Generative Adversarial Network (GAN). The key idea of Softmax GAN is to replace the classification loss in the original GAN with a softmax cross-e…
LoraHub: Efficient Cross-Task Generalization via Dynamic LoRA Composition
Chengsong Huang, Qian Liu, Bill Yuchen Lin +3
Low-rank adaptations (LoRA) are often employed to fine-tune large language models (LLMs) for new tasks. This paper investigates LoRA composability for cross-task generalization and…
SPIRAL: Self-Play on Zero-Sum Games Incentivizes Reasoning via Multi-Agent Multi-Turn Reinforcement Learning
Bo Liu, Leon Guertler, Simon Yu +9
Recent advances in reinforcement learning have shown that language models can develop sophisticated reasoning through training on tasks with verifiable rewards, but these approache…
The hidden spin-momentum locking and topological defects in unpolarized light fields
Peng Shi, Min Lin, Xinxin Gou +3
Electromagnetic waves characterized by intensity, phase, and polarization degrees of freedom are widely applied in data storage, encryption, and communications. However, these prop…
Thickness-dependent Dielectric Constant of Few-layer In2Se3 Nano-flakes
Di Wu, Alexander J. Pak, Yingnan Liu +10
The dielectric constant or relative permittivity of a dielectric material, which describes how the net electric field in the medium is reduced with respect to the external field, i…
CINO: A Chinese Minority Pre-trained Language Model
Ziqing Yang, Zihang Xu, Yiming Cui +4
Multilingual pre-trained language models have shown impressive performance on cross-lingual tasks. It greatly facilitates the applications of natural language processing on low-res…
Secrecy Capacity Bounds for Visible Light Communications With Signal-Dependent Noise
Jin-Yuan Wang, Xian-Tao Fu, Jun-Bo Wang +3
In physical-layer security, one of the most fundamental issues is the secrecy capacity. The objective of this paper is to determine the secrecy capacity for an indoor visible light…
Measuring vacancy-type defect density in monolayer semiconductors
Aleksandar Radic, Nick von Jeinsen, Vivian Perez +12
Two-dimensional (2D) materials have attracted wide-spread interest due to their unique and tunable properties. Their optoelectronic, mechanical, and thermal properties are greatly…
Improved Few-Shot Jailbreaking Can Circumvent Aligned Language Models and Their Defenses
Xiaosen Zheng, Tianyu Pang, Chao Du +3
Recently, Anil et al. (2024) show that many-shot (up to hundreds of) demonstrations can jailbreak state-of-the-art LLMs by exploiting their long-context capability. Nevertheless, i…
Lifelong Safety Alignment for Language Models
Haoyu Wang, Zeyu Qin, Yifei Zhao +4
LLMs have made impressive progress, but their growing capabilities also expose them to highly flexible jailbreaking attacks designed to bypass safety alignment. While many existing…
Bootstrapping Language Models with DPO Implicit Rewards
Changyu Chen, Zichen Liu, Chao Du +5
Human alignment in large language models (LLMs) is an active area of research. A recent groundbreaking work, direct preference optimization (DPO), has greatly simplified the proces…
Correntropy Induced L2 Graph for Robust Subspace Clustering
Canyi Lu, Jinhui Tang, Min Lin +3
In this paper, we study the robust subspace clustering problem, which aims to cluster the given possibly noisy data points into their underlying subspaces. A large pool of previous…
Balancing Pipeline Parallelism with Vocabulary Parallelism
Man Tsung Yeung, Penghui Qi, Min Lin +1
Pipeline parallelism is widely used to scale the training of transformer-based large language models, various works have been done to improve its throughput and memory footprint. I…
On the role of a priori knowledge in the optimization of quantum information processing
Ming Zhang, Min Lin, S. G. Schirmer +3
This paper explores the role of a priori knowledge in the optimization of quantum information processing by investigating optimum unambiguous discrimination problems for both the q…
Test-Time Backdoor Attacks on Multimodal Large Language Models
Dong Lu, Tianyu Pang, Chao Du +3
Backdoor attacks are commonly executed by contaminating training data, such that a trigger can activate predetermined harmful effects during the test phase. In this work, we presen…
PCSTracker: Long-Term Scene Flow Estimation for Point Cloud Sequences
Min Lin, Gangwei Xu, Xianqi Wang +2
Point cloud scene flow estimation is fundamental to long-term and fine-grained 3D motion analysis. However, existing methods are typically limited to pairwise settings and struggle…
LightTransfer: Your Long-Context LLM is Secretly a Hybrid Model with Effortless Adaptation
Xuan Zhang, Fengzhuo Zhang, Cunxiao Du +4
Scaling language models to handle longer contexts introduces substantial memory challenges due to the growing cost of key-value (KV) caches. Motivated by the efficiency gains of hy…
GEM: A Gym for Agentic LLMs
Zichen Liu, Anya Sims, Keyu Duan +16
The training paradigm for large language models (LLMs) is moving from static datasets to experience-based learning, where agents acquire skills via interacting with complex environ…
Variational Reasoning for Language Models
Xiangxin Zhou, Zichen Liu, Haonan Wang +5
We introduce a variational reasoning framework for language models that treats thinking traces as latent variables and optimizes them through variational inference. Starting from t…
Correlated Dirac eigenvalues around the transition temperature on lattices
Heng-Tong Ding, Wei-Ping Huang, Min Lin +3
We investigate the criticality of chiral phase transition manifested in the first and second order derivatives of Dirac eigenvalue spectrum with respect to light quark mass in (2+1…
MXNet: A Flexible and Efficient Machine Learning Library for Heterogeneous Distributed Systems
Tianqi Chen, Mu Li, Yutian Li +7
MXNet is a multi-language machine learning (ML) library to ease the development of ML algorithms, especially for deep neural networks. Embedded in the host language, it blends decl…
EchoVLA: Robotic Vision-Language-Action Model with Synergistic Declarative Memory for Mobile Manipulation
Min Lin, Xiwen Liang, Bingqian Lin +13
Recent progress in Vision-Language-Action (VLA) models has enabled embodied agents to interpret multimodal instructions and perform complex tasks. However, existing VLAs are mostly…
Mutual Information Regularized Offline Reinforcement Learning
Xiao Ma, Bingyi Kang, Zhongwen Xu +2
The major challenge of offline RL is the distribution shift that appears when out-of-distribution actions are queried, which makes the policy improvement direction biased by extrap…