Author name cluster

Haoran Xu

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

30 papers

2 author rows

AAAI Conference 2026 Conference Paper

BulletTime4D: Towards High Spatio-Temporal Resolution Dynamic Scene Rendering via Spike-Guided Stereo Vision

Yiqian Chang
Haoran Xu
Qinghong Ye
Jianing Li
Xuan Wang
Wei Zhang
Peixi Peng

High spatio‑temporal resolution novel‑view scene rendering is crucial for applications such as sports analysis and scientific experiments. However, existing Dynamic Scene Rendering (DSR) approaches typically rely on conventional RGB cameras with limited frame rates, making it difficult to achieve high spatio‑temporal resolution. In this paper, we present BulletTime4D, a high spatio‑temporal resolution DSR framework, which is the first trial to integrate a spike camera with binocular RGB cameras for dynamic scene reconstruction. Specifically, we first develop a hybrid camera prototype and build a real‑world dynamic scene reconstruction dataset. Then, BulletTime4D presents a multi‑timescale deformation representation by combining low‑frequency spatio‑temporal features with high‑frequency inter‑frame motion features. Finally, a rendering network is designed capable of projecting 4D Gaussians into the spike domain for spike rendering, and a cross‑domain supervision strategy is proposed to achieve high‑frame‑rate texture and color rendering. The results show that BulletTime4D outperforms state‑of‑the‑art methods on both simulated and real‑world datasets. In addition, BulletTime4D can synthesize 300 FPS novel‑view renderings using stereo RGB cameras at 30 FPS and a single spike camera.

PDF Details DOI

AAAI Conference 2026 Conference Paper

COVR: Collaborative Optimization of VLMs and RL Agent for Visual-Based Control

Canming Xia
Peixi Peng
Guang Tan
Zhan Su
Haoran Xu
Zhenxian Liu
Luntong Li

Visual reinforcement learning (RL) suffers from poor sample efficiency due to high-dimensional observations in complex tasks. While existing works have shown that vision-language models (VLMs) can assist RL, they often focus on knowledge distillation from the VLM to RL, overlooking the potential of RL-generated interaction data to enhance the VLM. To address this, we propose COVR, a collaborative optimization framework that enables the mutual enhancement of the VLM and RL policies. Specifically, COVR fine-tunes the VLM with RL-generated data to enhance the semantic reasoning ability consistent with the target task, and uses the enhanced VLM to further guide policy learning via action priors. To improve fine-tuning efficiency, we introduce two key modules: (1) an Exploration-Driven Dynamic Filter module that preserves valuable exploration samples using adaptive thresholds based on the degree of exploration, and (2) a Return-Aware Adaptive Loss Weight module that improves the stability of training by quantifying the inconsistency of sampling actions via return signals of RL. We further design a progressive fine-tuning strategy to reduce resource consumption. Extensive experiments show that COVR achieves strong performance across various challenging visual control tasks.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Diffusion Distillation with Direct Preference Optimization for Efficient 3D LiDAR Scene Completion

An Zhao
Shengyuan Zhang
Zejian Li
Ling Yang
Pei Chen
Jiale Wu
Haoran Xu
AnYang Wei

The slow sampling speed of diffusion models hinders their application in 3D LiDAR scene completion. To address this, we propose Distillation-DPO, a novel framework that accelerates sampling through score distillation while simultaneously enhancing generation quality via preference alignment. Distillation-DPO follows a three-step procedure. First, the student model generates paired completion scenes with different initial noises. Second, using LiDAR scene evaluation metrics as preference, we construct winning and losing sample pairs. Third, as our core innovation, Distillation-DPO optimizes the student model by exploiting the difference in score functions between the teacher and student models on the paired completion scenes. This operation performs variational score distillation of the student model but simultaneously encourages the distilled student to prefer the winning samples over the losing ones. Extensive experiments demonstrate that Distillation-DPO achieves higher-quality scene completion than state-of-the-art diffusion models, while accelerating sampling by over 5-fold. To our knowledge, our work is the first to integrate the preference learning principle of DPO into the distillation of diffusion models, offering a new framework of preference-aligned distillation.

PDF Details DOI

AAAI Conference 2026 Conference Paper

READ: Real-time and Efficient Asynchronous Diffusion for Audio-driven Talking Head Generation

Haotian Wang
Yuzhe Weng
Jun Du
Haoran Xu
Xiaoyan Wu
Shan He
Bing Yin
Cong Liu

The introduction of diffusion models has brought significant advances to the field of audio-driven talking head generation. However, the extremely slow inference speed severely limits the practical implementation of diffusion-based talking head generation models. In this study, we propose READ, a real-time diffusion-transformer-based talking head generation framework. Our approach first learns a spatiotemporal highly compressed video latent space via a temporal VAE, significantly reducing the token count to accelerate generation. To achieve better audio-visual alignment within this compressed latent space, a pre-trained Speech Autoencoder (SpeechAE) is proposed to generate temporally compressed speech latent codes corresponding to the video latent space. These latent representations are then modeled by a carefully designed Audio-to-Video Diffusion Transformer (A2V-DiT) backbone for efficient talking head synthesis. Furthermore, to ensure temporal consistency and accelerated inference in extended generation, we propose a novel asynchronous noise scheduler (ANS) for both the training and inference processes of our framework. The ANS leverages asynchronous add-noise and asynchronous motion-guided generation in the latent space, ensuring consistency in generated video clips. Experimental results demonstrate that READ outperforms state-of-the-art methods by generating competitive talking head videos with significantly reduced runtime, achieving an optimal balance between quality and speed while maintaining robust metric stability in long-time generation.

PDF Details DOI

AAAI Conference 2026 Conference Paper

SNN-Driven Event-Based Flow and Rotation Estimation with SO(3) Refinement

Ruimin Sun
Haoran Xu
De Ma

Spiking Neural Networks (SNNs) offer a promising direction for energy-efficient event-based vision by leveraging sparse, temporally precise spikes. We propose a directly trained, fully spiking model for optical flow estimation, featuring a novel Spike GRU and membrane potential carryover for improved temporal modeling. On the DSEC-Flow benchmark, our model achieves competitive accuracy while reducing energy consumption by 42.88× over EV-FlowNet and 38× over TIDNet. Building on the predicted motion field, we infer camera rotation and, to the best of our knowledge, are the first to construct panoramic event images from SNN-based flow. We further introduce an optional unsupervised SO(3) refinement step that improves rotation accuracy by maximizing panorama consistency—without IMU or pose supervision. Our results achieve comparable visual quality to CMax-SLAM, showing that SNNs can enable fast and high-level spatial perception using only event-based input.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Towards Affordance-Aware Robotic Dexterous Grasping with Human-like Priors

Haoyu Zhao
Linghao Zhuang
Xingyue Zhao
Cheng Zeng
Haoran Xu
Yuming Jiang
Jun CEN
Kexiang Wang

A dexterous hand capable of generalizable grasping objects is fundamental for the development of general-purpose embodied AI. However, previous methods focus narrowly on low-level grasp stability metrics, neglecting affordance-aware positioning and human-like poses which are crucial for downstream manipulation. To address these limitations, we propose AffordDex, a novel framework with two-stage training that learns a universal grasping policy with an inherent understanding of both motion priors and object affordances. In the first stage, a trajectory imitator is pre-trained on a large corpus of human hand motions to instill a strong prior for natural movement. In the second stage, a residual module is trained to adapt these general human-like motions to specific object instances. This refinement is critically guided by two components: our Negative Affordance-aware Segmentation (NAA) module, which identifies functionally inappropriate contact regions, and a privileged teacher-student distillation process that ensures the final vision-based policy is highly successful. Extensive experiments demonstrate that AffordDex not only achieves universal dexterous grasping but also remains remarkably human-like in posture and functionally appropriate in contact location. As a result, AffordDex significantly outperforms state-of-the-art baselines across seen objects, unseen instances, and even entirely novel categories.

PDF Details DOI

ICLR Conference 2025 Conference Paper

Adapters for Altering LLM Vocabularies: What Languages Benefit the Most?

HyoJung Han
Akiko Eriguchi
Haoran Xu
Hieu Hoang
Marine Carpuat
Huda Khayrallah

Vocabulary adaptation, which integrates new vocabulary into pre-trained language models, enables expansion to new languages and mitigates token over-fragmentation. However, existing approaches are limited by their reliance on heuristics or external embeddings. We propose VocADT, a novel method for vocabulary adaptation using adapter modules that are trained to learn the optimal linear combination of existing embeddings while keeping the model’s weights fixed. VocADT offers a flexible and scalable solution without depending on external resources or language constraints. Across 11 languages—with diverse scripts, resource availability, and fragmentation—we demonstrate that VocADT outperforms the original Mistral model (Jiang et al., 2023) and other baselines across various multilingual tasks including natural language understanding and machine translation. We find that Latin-script languages and highly fragmented languages benefit the most from vocabulary adaptation. We further fine-tune the adapted model on the generative task of machine translation and find that vocabulary adaptation is still beneficial after fine-tuning and that VocADT is the most effective.

Details

NeurIPS Conference 2025 Conference Paper

Decoder-Hybrid-Decoder Architecture for Efficient Reasoning with Long Generation

Liliang Ren
Congcong Chen
Haoran Xu
Young Jin Kim
Adam Atkinson
Zheng Zhan
Jiankai Sun
Baolin Peng

Recent advances in language modeling have demonstrated the effectiveness of State Space Models (SSMs) for efficient sequence modeling. While hybrid architectures such as Samba and the decoder-decoder architecture, YOCO, have shown promising performance gains over Transformers, prior works have not investigated the efficiency potential of representation sharing between SSM layers. In this paper, we introduce the Gated Memory Unit (GMU), a simple yet effective mechanism for efficient memory sharing across layers. We apply it to create SambaY, a decoder-hybrid-decoder architecture that incorporates GMUs in the cross-decoder to share memory readout states from a Samba-based self-decoder. SambaY significantly enhances decoding efficiency, preserves linear pre-filling time complexity, and boosts long-context performance, all while eliminating the need for explicit positional encoding. Through extensive scaling experiments, we demonstrate that our model exhibits a significantly lower irreducible loss compared to a strong YOCO baseline, indicating superior performance scalability under large-scale compute regimes. Our largest model enhanced with Differential Attention, Phi4-mini-Flash-Reasoning, achieves significantly better performance than Phi4-mini-Reasoning on reasoning tasks such as Math500, AIME24/25, and GPQA Diamond without any reinforcement learning, while delivering up to 10× higher decoding throughput on 2K-length prompts with 32K generation length under the vLLM inference framework. We release our training codebase on open-source data at https: //github. com/microsoft/ArchScale.

PDF Details

ICLR Conference 2025 Conference Paper

Dynamical Diffusion: Learning Temporal Dynamics with Diffusion Models

Xingzhuo Guo
Yu Zhang
Baixu Chen
Haoran Xu
Jianmin Wang 0001
Mingsheng Long

Diffusion models have emerged as powerful generative frameworks by progressively adding noise to data through a forward process and then reversing this process to generate realistic samples. While these models have achieved strong performance across various tasks and modalities, their application to temporal predictive learning remains underexplored. Existing approaches treat predictive learning as a conditional generation problem, but often fail to fully exploit the temporal dynamics inherent in the data, leading to challenges in generating temporally coherent sequences. To address this, we introduce Dynamical Diffusion (DyDiff), a theoretically sound framework that incorporates temporally aware forward and reverse processes. Dynamical Diffusion explicitly models temporal transitions at each diffusion step, establishing dependencies on preceding states to better capture temporal dynamics. Through the reparameterization trick, Dynamical Diffusion achieves efficient training and inference similar to any standard diffusion model. Extensive experiments across scientific spatiotemporal forecasting, video prediction, and time series forecasting demonstrate that Dynamical Diffusion consistently improves performance in temporal predictive tasks, filling a crucial gap in existing methodologies. Code is available at this repository: https://github.com/thuml/dynamical-diffusion.

Details

AAAI Conference 2025 Conference Paper

Exploiting Continuous Motion Clues for Vision-Based Occupancy Prediction

Haoran Xu
Peixi Peng
Xinyi Zhang
Guang Tan
Yaokun Li
Shuaixian Wang
Luntong Li

Occupancy networks aim to reconstruct the surroundings with occupied semantic voxels. However, frequent object occlusions often occur in dynamic real-world scenarios, which cannot be captured by independent frames. Most existing occupancy networks generate results without explicitly considering past occupancy states and continuous visual changes over time, limiting their temporal accuracy. We tackle it by treating the task from a new continuous updating perspective, which considers historical data and continuous motion clues. We propose a new approach termed Continuous Motion clue exploitation for Occupancy Prediction (CMOP), which incorporates three key designs: (i) Propagator: which forecasts future occupancy states based on historical data; (ii) Tracker: which updates the occupancy on a per-frame basis using dynamic visual motion information; and (iii) Fuser: which aggregates results from the Propagator and Tracker into more robust and accurate occupancy results. Experiments on several benchmarks demonstrate that CMOP outperforms state-of-the-art baselines.

PDF Details DOI

AAAI Conference 2025 Conference Paper

Federated Learning with Sample-level Client Drift Mitigation

Haoran Xu
Jiaze Li
Wanyi Wu
Hao Ren

Federated Learning (FL) suffers from severe performance degradation due to the data heterogeneity among clients. Existing works reveal that the fundamental reason is that data heterogeneity can cause client drift where the local model update deviates from the global one, and thus they usually tackle this problem from the perspective of calibrating the obtained local update. Despite effectiveness, existing methods substantially lack a deep understanding of how heterogeneous data samples contribute to the formation of client drift. In this paper, we bridge this gap by identifying that the drift can be viewed as a cumulative manifestation of biases present in all local samples and the bias between samples is different. Besides, the bias dynamically changes as the FL training progresses. Motivated by this, we propose FedBSS that first mitigates the heterogeneity issue in a sample-level manner, orthogonal to existing methods. Specifically, the core idea of our method is to adopt a bias-aware sample selection scheme that dynamically selects the samples from small biases to large epoch by epoch to train progressively the local model in each round. In order to ensure the stability of training, we set the diversified knowledge acquisition stage as the warm-up stage to avoid the local optimality caused by knowledge deviation in the early stage of the model. Evaluation results show that FedBSS outperforms state-of-the-art baselines. In addition, we also achieved effective results on feature distribution skew and noise label dataset setting, which proves that FedBSS can not only reduce heterogeneity, but also has scalability and robustness.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

Information-Theoretic Reward Decomposition for Generalizable RLHF

Liyuan Mao
Haoran Xu
Amy Zhang
Weinan Zhang
Chenjia Bai

Obtaining a generalizable reward model is crucial in Reinforcement Learning from Human Feedback (RLHF) as it enables correctly evaluating unseen prompt-response pairs. However, existing reward models lack this ability, as they are typically trained by increasing the reward gap between chosen and rejected responses, while overlooking the prompts that the responses are conditioned on. Consequently, when the trained reward model is evaluated on prompt-response pairs that lie outside the data distribution, neglecting the effect of prompts may result in poor generalization of the reward model. To address this issue, we decompose the reward value into two independent components: prompt-free reward and prompt-related reward. Prompt-free reward represents the evaluation that is determined only by responses, while the prompt-related reward reflects the reward that derives from both the prompt and the response. We extract these two components from an information theoretical perspective, which requires no extra models. . Subsequently, we propose a new reward learning algorithm by prioritizing data samples based on their prompt-free reward values. Through toy examples, we demonstrate that the extracted prompt-free and prompt-related rewards effectively characterize two parts of the reward model. Further, standard evaluations show that our method improves both the alignment performance and the generalization capability of the reward model.

PDF Details

AAAI Conference 2025 Conference Paper

MVReward: Better Aligning and Evaluating Multi-View Diffusion Models with Human Preferences

Weitao Wang
Haoran Xu
Yuxiao Yang
Zhifang Liu
Jun Meng
Haoqian Wang

Recent years have witnessed remarkable progress in 3D content generation. However, corresponding evaluation methods struggle to keep pace. Automatic approaches have proven challenging to align with human preferences, and the mixed comparison of text- and image-driven methods often leads to unfair evaluations. In this paper, we present a comprehensive framework to better align and evaluate multi-view diffusion models with human preferences. To begin with, we first collect and filter a standardized image prompt set from DALL·E and Objaverse, which we then use to generate multi-view assets with several multi-view diffusion models. Through a systematic ranking pipeline on these assets, we obtain a human annotation dataset with 16k expert pairwise comparisons and train a reward model, coined MVReward, to effectively encode human preferences. With MVReward, image-driven 3D methods can be evaluated against each other in a more fair and transparent manner. Building on this, we further propose Multi-View Preference Learning (MVP), a plug-and-play multi-view diffusion tuning strategy. Extensive experiments demonstrate that MVReward can serve as a reliable metric and MVP consistently enhances the alignment of multi-view diffusion models with human preferences.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

Spike4DGS: Towards High-Speed Dynamic Scene Rendering with 4D Gaussian Splatting via a Spike Camera Array

Qinghong Ye
Yiqian Chang
Jianing Li
Haoran Xu
Xuan Wang
Wei Zhang
Yonghong Tian
Peixi Peng

Spike camera with high temporal resolution offers a new perspective on high-speed dynamic scene rendering. Most existing rendering methods rely on Neural Radiance Fields (NeRF) or 3D Gaussian Splatting (3DGS) for static scenes using a monocular spike camera. However, these methods struggle with dynamic motion, while a single camera suffers from limited spatial coverage, making it challenging to reconstruct fine details in high-speed scenes. To address these problems, we propose Spike4DGS, the first high-speed dynamic scene rendering framework with 4D Gaussian Splatting using spike camera arrays. Technically, we first build a multi-view spike camera array to validate our solution, then establish both synthetic and real-world multi-view spike-based reconstruction datasets. Then, we design a multi-view spike-based dense initialization module that obtains dense point clouds and camera poses from continuous spike streams. Finally, we propose a spike-pixel synergy constraint supervision to optimize Spike4DGS, incorporating both rendered image quality loss and dynamic spatiotemporal spike loss. The results show that our Spike4DGS outperforms state-of-the-art methods in terms of novel view rendering quality on both synthetic and real-world datasets. More details are available at https: //github. com/Qinghongye/Spike4DGS.

PDF Details

ICLR Conference 2025 Conference Paper

The Belief State Transformer

Edward S. Hu
Kwangjun Ahn
Qinghua Liu
Haoran Xu
Manan Tomar
Ada Langford
Dinesh Jayaraman
Alex Lamb

We introduce the "Belief State Transformer", a next-token predictor that takes both a prefix and suffix as inputs, with a novel objective of predicting both the next token for the prefix and the previous token for the suffix. The Belief State Transformer effectively learns to solve challenging problems that conventional forward-only transformers struggle with, in a domain-independent fashion. Key to this success is learning a compact belief state that captures all relevant information necessary for accurate predictions. Empirical ablations show that each component of the model is essential in difficult scenarios where standard Transformers fall short. For the task of story writing with known prefixes and suffixes, our approach outperforms the Fill-in-the-Middle method for reaching known goals and demonstrates improved performance even when the goals are unknown. Altogether, the Belief State Transformer enables more efficient goal-conditioned decoding, better test-time inference, and high-quality text representations on small scale problems. Website: https://edwhu.github.io/bst-website

Details

NeurIPS Conference 2025 Conference Paper

Uni-RL: Unifying Online and Offline RL via Implicit Value Regularization

Haoran Xu
Liyuan Mao
Hui Jin
Weinan Zhang
Xianyuan Zhan
Amy Zhang

The practical use of reinforcement learning (RL) requires handling diverse settings, including online, offline, and offline-to-online learning. Instead of developing separate algorithms for each setting, we propose Uni-RL, a unified model-free RL framework that addresses all these scenarios within a single formulation. Uni-RL builds on the Implicit Value Regularization (IVR) framework and generalizes its dataset behavior constraint to the constraint w. r. t a reference policy, yielding an unified value learning objective for general settings. The reference policy is chosen to be the target policy in the online setting and the behavior policy in the offline setting. Using an iteratively refined behavior policy solves the over-constrained problem of directly applying IVR in the online setting, it provides an implicit trust-region style update through the value function while being off-policy. Uni-RL also introduces an unified policy extraction objective that estimates in-sample policy gradient using only actions from the reference policy. This supports various policy classes and theoretically guaranntees less value estimation error and larger performance improvement over the reference policy. We evaluate Uni-RL on a range of standard RL benchmarks across online, offline, and offline-to-online settings. In online RL, Uni-RL achieves higher sample efficiency than both off-policy methods without trust-region updates and on-policy methods with trust-region updates. In offline RL, Uni-RL retains the benefits of in-sample learning while outperforming IVR through better policy extraction. In offline-to-online RL, Uni-RL beats both constraint-based methods and unconstrained approaches by effectively balancing stability and adaptability.

PDF Details

ICLR Conference 2025 Conference Paper

X-ALMA: Plug & Play Modules and Adaptive Rejection for Quality Translation at Scale

Haoran Xu
Kenton Murray
Philipp Koehn
Hieu Hoang
Akiko Eriguchi
Huda Khayrallah

Large language models (LLMs) have achieved remarkable success across various NLP tasks with a focus on English due to English-centric pre-training and limited multilingual data. In this work, we focus on the problem of translation, and while some multilingual LLMs claim to support for hundreds of languages, models often fail to provide high-quality responses for mid- and low-resource languages, leading to imbalanced performance heavily skewed in favor of high-resource languages. We introduce **X-ALMA**, a model designed to ensure top-tier performance across 50 diverse languages, regardless of their resource levels. X-ALMA surpasses state-of-the-art open-source multilingual LLMs, such as Aya-101 and Aya-23, in every single translation direction on the FLORES-200 and WMT'23 test datasets according to COMET-22. This is achieved by plug-and-play language-specific module architecture to prevent language conflicts during training and a carefully designed training regimen with novel optimization methods to maximize the translation performance. After the final stage of training regimen, our proposed **A**daptive **R**ejection **P**reference **O**ptimization (**ARPO**) surpasses existing preference optimization methods in translation tasks.

Details

ICLR Conference 2024 Conference Paper

A Paradigm Shift in Machine Translation: Boosting Translation Performance of Large Language Models

Haoran Xu
Young Jin Kim 0001
Amr Sharaf
Hany Hassan

Generative Large Language Models (LLMs) have achieved remarkable advancements in various NLP tasks. However, these advances have not been reflected in the translation task, especially those with moderate model sizes (i.e., 7B or 13B parameters), which still lag behind conventional supervised encoder-decoder translation models. Previous studies have attempted to improve the translation capabilities of these LLMs, but their gains have been limited. In this study, we propose a novel fine-tuning approach for LLMs that is specifically designed for the translation task, eliminating the need for the abundant parallel data that traditional translation models usually depend on. Our approach consists of two fine-tuning stages: initial fine-tuning on monolingual data followed by subsequent fine-tuning on a small set of high-quality parallel data. We introduce the LLM developed through this strategy as **A**dvanced **L**anguage **M**odel-based tr**A**nslator (**ALMA**). Based on LLaMA-2 as our underlying model, our results show that the model can achieve an average improvement of more than 12 BLEU and 12 COMET over its zero-shot performance across 10 translation directions from the WMT'21 (2 directions) and WMT'22 (8 directions) test datasets. The performance is significantly better than all prior work and even superior to the NLLB-54B model \citep{nllb} and GPT-3.5-text-davinci-003, with only 7B or 13B parameters. This method establishes the foundation for a novel training paradigm in machine translation.

Details

ICML Conference 2024 Conference Paper

Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation

Haoran Xu
Amr Sharaf
Yunmo Chen
Weiting Tan
Lingfeng Shen
Benjamin Van Durme
Kenton Murray
Young Jin Kim 0001

Moderate-sized large language models (LLMs) – those with 7B or 13B parameters – exhibit promising machine translation (MT) performance. However, they do not match the performance of state-of-the-art conventional encoder-decoder translation models or larger-scale LLMs such as GPT-4. In this study, we bridge this performance gap. We first assess the shortcomings of supervised fine-tuning for LLMs in the MT task, emphasizing the quality issues present in the reference data, despite being human-generated. Then, in contrast to supervised fine-tuning which mimics reference translations, we introduce Contrastive Preference Optimization (CPO), a novel approach that trains models to avoid generating adequate but not perfect translations. Applying CPO to ALMA models with only 22K parallel sentences and 0. 1% parameters yields significant improvements. The resulting model, called ALMA-R, can match or exceed the performance of the WMT competition winners and GPT-4 on WMT’21, WMT’22 and WMT’23 test datasets.

Details

NeurIPS Conference 2024 Conference Paper

Diffusion-DICE: In-Sample Diffusion Guidance for Offline Reinforcement Learning

Liyuan Mao
Haoran Xu
Xianyuan Zhan
Weinan Zhang
Amy Zhang

One important property of DIstribution Correction Estimation (DICE) methods is that the solution is the optimal stationary distribution ratio between the optimized and data collection policy. In this work, we show that DICE-based methods can be viewed as a transformation from the behavior distribution to the optimal policy distribution. Based on this, we propose a novel approach, Diffusion-DICE, that directly performs this transformation using diffusion models. We find that the optimal policy's score function can be decomposed into two terms: the behavior policy's score function and the gradient of a guidance term which depends on the optimal distribution ratio. The first term can be obtained from a diffusion model trained on the dataset and we propose an in-sample learning objective to learn the second term. Due to the multi-modality contained in the optimal policy distribution, the transformation in Diffusion-DICE may guide towards those local-optimal modes. We thus generate a few candidate actions and carefully select from them to achieve global-optimum. Different from all other diffusion-based offline RL methods, the \textit{guide-then-select} paradigm in Diffusion-DICE only uses in-sample actions for training and brings minimal error exploitation in the value function. We use a didatic toycase example to show how previous diffusion-based methods fail to generate optimal actions due to leveraging these errors and how Diffusion-DICE successfully avoid that. We then conduct extensive experiments on benchmark datasets to show the strong performance of Diffusion-DICE.

PDF Details DOI

ICLR Conference 2024 Conference Paper

Error Norm Truncation: Robust Training in the Presence of Data Noise for Text Generation Models

Tianjian Li
Haoran Xu
Philipp Koehn
Daniel Khashabi
Kenton Murray

Text generation models are notoriously vulnerable to errors in the training data. With the wide-spread availability of massive amounts of web-crawled data becoming more commonplace, how can we enhance the robustness of models trained on a massive amount of noisy web-crawled text? In our work, we propose Error Norm Truncation (ENT), a robust enhancement method to the standard training objective that truncates noisy data. Compared to methods that only uses the negative log-likelihood loss to estimate data quality, our method provides a more accurate estimation by considering the distribution of non-target tokens, which is often overlooked by previous work. Through comprehensive experiments across language modeling, machine translation, and text summarization, we show that equipping text generation models with ENT improves generation quality over standard training and previous soft and hard truncation methods. Furthermore, we show that our method improves the robustness of models against two of the most detrimental types of noise in machine translation, resulting in an increase of more than 2 BLEU points over the MLE baseline when up to 50\% of noise is added to the data.

Details

ICLR Conference 2024 Conference Paper

FedCDA: Federated Learning with Cross-rounds Divergence-aware Aggregation

Haozhao Wang
Haoran Xu
Yichen Li 0006
Yuan Xu
Ruixuan Li 0001
Tianwei Zhang 0004

In Federated Learning (FL), model aggregation is pivotal. It involves a global server iteratively aggregating client local trained models in successive rounds without accessing private data. Traditional methods typically aggregate the local models from the current round alone. However, due to the statistical heterogeneity across clients, the local models from different clients may be greatly diverse, making the obtained global model incapable of maintaining the specific knowledge of each local model. In this paper, we introduce a novel method, FedCDA, which selectively aggregates cross-round local models, decreasing discrepancies between the global model and local models. The principle behind FedCDA is that due to the different global model parameters received in different rounds and the non-convexity of deep neural networks, the local models from each client may converge to different local optima across rounds. Therefore, for each client, we select a local model from its several recent local models obtained in multiple rounds, where the local model is selected by minimizing its divergence from the local models of other clients. This ensures the aggregated global model remains close to all selected local models to maintain their data knowledge. Extensive experiments conducted on various models and datasets reveal our approach outperforms state-of-the-art aggregation methods.

Details

NeurIPS Conference 2023 Conference Paper

Hierarchical Adaptive Value Estimation for Multi-modal Visual Reinforcement Learning

Yangru Huang
Peixi Peng
Yifan Zhao
Haoran Xu
Mengyue Geng
Yonghong Tian

Integrating RGB frames with alternative modality inputs is gaining increasing traction in many vision-based reinforcement learning (RL) applications. Existing multi-modal vision-based RL methods usually follow a Global Value Estimation (GVE) pipeline, which uses a fused modality feature to obtain a unified global environmental description. However, such a feature-level fusion paradigm with a single critic may fall short in policy learning as it tends to overlook the distinct values of each modality. To remedy this, this paper proposes a Local modality-customized Value Estimation (LVE) paradigm, which dynamically estimates the contribution and adjusts the importance weight of each modality from a value-level perspective. Furthermore, a task-contextual re-fusion process is developed to achieve a task-level re-balance of estimations from both feature and value levels. To this end, a Hierarchical Adaptive Value Estimation (HAVE) framework is formed, which adaptively coordinates the contributions of individual modalities as well as their collective efficacy. Agents trained by HAVE are able to exploit the unique characteristics of various modalities while capturing their intricate interactions, achieving substantially improved performance. We specifically highlight the potency of our approach within the challenging landscape of autonomous driving, utilizing the CARLA benchmark with neuromorphic event and depth data to demonstrate HAVE's capability and the effectiveness of its distinct components.

PDF Details

YNIMG Journal 2023 Journal Article

Mesoscale functional connectivity in macaque visual areas

Xingya Cai
Haoran Xu
Chao Han
Peichao Li
Jiayu Wang
Rui Zhang
Rendong Tang
Chen Fang

Studies of resting-state functional connectivity (rsFC) have provided rich insights into the structures and functions of the human brain. However, most rsFC studies have focused on large-scale brain connectivity. To explore rsFC at a finer scale, we used intrinsic signal optical imaging to image the ongoing activity of the anesthetized macaque visual cortex. Differential signals from functional domains were used to quantify network-specific fluctuations. In 30-60 min resting-state imaging, a series of coherent activation patterns were observed in all three visual areas we examined (V1, V2, and V4). These patterns matched the known functional maps (ocular dominance, orientation, color) obtained in visual stimulation conditions. These functional connectivity (FC) networks fluctuated independently over time and exhibited similar temporal characteristics. Coherent fluctuations, however, were observed from orientation FC networks in different areas and even across two hemispheres. Thus, FC in the macaque visual cortex was fully mapped both on a fine scale and over a long range. Hemodynamic signals can be used to explore mesoscale rsFC in a submillimeter resolution.

Details DOI

NeurIPS Conference 2023 Conference Paper

Offline Multi-Agent Reinforcement Learning with Implicit Global-to-Local Value Regularization

Xiangsen Wang
Haoran Xu
Yinan Zheng
Xianyuan Zhan

Offline reinforcement learning (RL) has received considerable attention in recent years due to its attractive capability of learning policies from offline datasets without environmental interactions. Despite some success in the single-agent setting, offline multi-agent RL (MARL) remains to be a challenge. The large joint state-action space and the coupled multi-agent behaviors pose extra complexities for offline policy optimization. Most existing offline MARL studies simply apply offline data-related regularizations on individual agents, without fully considering the multi-agent system at the global level. In this work, we present OMIGA, a new offline multi-agent RL algorithm with implicit global-to-local value regularization. OMIGA provides a principled framework to convert global-level value regularization into equivalent implicit local value regularizations and simultaneously enables in-sample learning, thus elegantly bridging multi-agent value decomposition and policy learning with offline regularizations. Based on comprehensive experiments on the offline multi-agent MuJoCo and StarCraft II micro-management tasks, we show that OMIGA achieves superior performance over the state-of-the-art offline MARL methods in almost all tasks.

PDF Details

NeurIPS Conference 2022 Conference Paper

A Policy-Guided Imitation Approach for Offline Reinforcement Learning

Haoran Xu
Li Jiang
Li Jianxiong
Xianyuan Zhan

Offline reinforcement learning (RL) methods can generally be categorized into two types: RL-based and Imitation-based. RL-based methods could in principle enjoy out-of-distribution generalization but suffer from erroneous off-policy evaluation. Imitation-based methods avoid off-policy evaluation but are too conservative to surpass the dataset. In this study, we propose an alternative approach, inheriting the training stability of imitation-style methods while still allowing logical out-of-distribution generalization. We decompose the conventional reward-maximizing policy in offline RL into a guide-policy and an execute-policy. During training, the guide-poicy and execute-policy are learned using only data from the dataset, in a supervised and decoupled manner. During evaluation, the guide-policy guides the execute-policy by telling where it should go so that the reward can be maximized, serving as the \textit{Prophet}. By doing so, our algorithm allows \textit{state-compositionality} from the dataset, rather than \textit{action-compositionality} conducted in prior imitation-style methods. We dumb this new approach Policy-guided Offline RL (\texttt{POR}). \texttt{POR} demonstrates the state-of-the-art performance on D4RL, a standard benchmark for offline RL. We also highlight the benefits of \texttt{POR} in terms of improving with supplementary suboptimal data and easily adapting to new tasks by only changing the guide-poicy.

PDF Details

AAAI Conference 2022 Conference Paper

Constraints Penalized Q-learning for Safe Offline Reinforcement Learning

Haoran Xu
Xianyuan Zhan
Xiangyu Zhu

We study the problem of safe offline reinforcement learning (RL), the goal is to learn a policy that maximizes long-term reward while satisfying safety constraints given only offline data, without further interaction with the environment. This problem is more appealing for real world RL applications, in which data collection is costly or dangerous. Enforcing constraint satisfaction is non-trivial, especially in offline settings, as there is a potential large discrepancy between the policy distribution and the data distribution, causing errors in estimating the value of safety constraints. We show that naïve approaches that combine techniques from safe RL and offline RL can only learn sub-optimal solutions. We thus develop a simple yet effective algorithm, Constraints Penalized Q-Learning (CPQ), to solve the problem. Our method admits the use of data generated by mixed behavior policies. We present a theoretical analysis and demonstrate empirically that our approach can learn robustly across a variety of benchmark control tasks, outperforming several baselines.

PDF Details

AAAI Conference 2022 Conference Paper

DeepThermal: Combustion Optimization for Thermal Power Generating Units Using Offline Reinforcement Learning

Xianyuan Zhan
Haoran Xu
Yue Zhang
Xiangyu Zhu
Honglei Yin
Yu Zheng

Optimizing the combustion efficiency of a thermal power generating unit (TPGU) is a highly challenging and critical task in the energy industry. We develop a new data-driven AI system, namely DeepThermal, to optimize the combustion control strategy for TPGUs. At its core, is a new model-based offline reinforcement learning (RL) framework, called MORE, which leverages historical operational data of a TGPU to solve a highly complex constrained Markov decision process problem via purely offline training. In DeepThermal, we first learn a data-driven combustion process simulator from the offline dataset. The RL agent of MORE is then trained by combining real historical data as well as carefully filtered and processed simulation data through a novel restrictive exploration scheme. DeepThermal has been successfully deployed in four large coal-fired thermal power plants in China. Real-world experiments show that DeepThermal effectively improves the combustion efficiency of TPGUs. We also report the superior performance of MORE by comparing with the state-of-the-art algorithms on the standard offline RL benchmarks.

PDF Details

IJCAI Conference 2022 Conference Paper

Model-Based Offline Planning with Trajectory Pruning

Xianyuan Zhan
Xiangyu Zhu
Haoran Xu

The recent offline reinforcement learning (RL) studies have achieved much progress to make RL usable in real-world systems by learning policies from pre-collected datasets without environment interaction. Unfortunately, existing offline RL methods still face many practical challenges in real-world system control tasks, such as computational restriction during agent training and the requirement of extra control flexibility. The model-based planning framework provides an attractive alternative. However, most model-based planning algorithms are not designed for offline settings. Simply combining the ingredients of offline RL with existing methods either provides over-restrictive planning or leads to inferior performance. We propose a new light-weighted model-based offline planning framework, namely MOPP, which tackles the dilemma between the restrictions of offline learning and high-performance planning. MOPP encourages more aggressive trajectory rollout guided by the behavior policy learned from data, and prunes out problematic trajectories to avoid potential out-of-distribution samples. Experimental results show that MOPP provides competitive performance compared with existing model-based offline planning and RL approaches.

PDF Details DOI

AAAI Conference 2021 Conference Paper

Robust Spatio-Temporal Purchase Prediction via Deep Meta Learning

Huiling Qin
Songyu Ke
Xiaodu Yang
Haoran Xu
Xianyuan Zhan
Yu Zheng

Purchase prediction is an essential task in both online and offline retail industry, especially during major shopping festivals, when strong promotion boosts consumption dramatically. It is important for merchants to forecast such surge of sales and have better preparation. This is a challenging problem, as the purchase patterns during shopping festivals are significantly different from usual cases and also rare in historical data. Most existing methods fail at this problem due to the extremely scarce data samples as well as the inability to capture the complex macroscopic spatio-temporal dependencies in a city. To address this problem, we propose the Spatio-Temporal Meta-learning Prediction (STMP) model for purchase prediction during shopping festivals. STMP is a meta-learning based spatio-temporal multi-task deep generative model. It adopts a meta-learning framework with few-shot learning capability to capture both spatial and temporal data representations. A generative component then uses the extracted spatio-temporal representation and input data to infer the prediction results. Extensive experiments demonstrate the meta-learning generalization ability of STMP. STMP outperforms baselines in all cases, which shows the effectiveness of our model.

PDF Details