papers

Publications (57)

cs.CV2026

Qwen-Image-Bench: From Generation to Creation in Text-to-Image Evaluation

Niantong Li, Guangzheng Hu, Weixu Qiao +35

Text-to-Image generation has evolved from basic image synthesis into a frequently used core capability in professional creative workflows, where simple text-image alignment can no…

cs.DC2026

SwiftCache: Efficient LLM Serving for Multi-turn Conversations with Heterogeneous KV Cache Sharing

Jianmin Hu, Minxian Xu, Sa Wang +5

Multi-turn conversation is a fundamental scenario in LLM applications, widely used in chatbots and AI agents. As the conversation evolves, historical tokens accumulate continuously…

cs.LG2025

Part II: ROLL Flash -- Accelerating RLVR and Agentic Training with Asynchrony

Han Lu, Zichen Liu, Shaopan Xiong +19

Synchronous Reinforcement Learning (RL) post-training has emerged as a crucial step for enhancing Large Language Models (LLMs) with diverse capabilities. However, many systems desi…

cs.LG2021

M6-T: Exploring Sparse Expert Models and Beyond

An Yang, Junyang Lin, Rui Men +12

Mixture-of-Experts (MoE) models can achieve promising results with outrageous large amount of parameters but constant computation cost, and thus it has become a trend in model scal…

cs.LG2026

Semantic DLM+: Improving Diffusion Language Models through Bias-variance Trade-off in Transition Kernel Design

Keyue Jiang, Yuxiang Wang, Yanan Zhao +7

Diffusion Language Models (DLMs) have demonstrated strong scaling capacity as alternatives to autoregressive language models. However, their performance is highly sensitive to the…

cs.AI2026

Let It Flow: Agentic Crafting on Rock and Roll, Building the ROME Model within an Open Agentic Learning Ecosystem

Weixun Wang, XiaoXiao Xu, Wanhe An +86

Agentic crafting requires LLMs to operate in real-world environments over multiple turns by taking actions, observing outcomes, and iteratively refining artifacts. Despite its impo…

cs.CL2024

D-CPT Law: Domain-specific Continual Pre-Training Scaling Law for Large Language Models

Haoran Que, Jiaheng Liu, Ge Zhang +13

Continual Pre-Training (CPT) on Large Language Models (LLMs) has been widely used to expand the model's fundamental understanding of specific downstream domains (e.g., math and cod…

cs.CL2026

Are the Financial Reasoning from LLMs Credible? A Real World Test over Long-Horizon Statements

Xinke Tong, Xuanming Zhang, Tianyi Tang +10

Do Large Language Models (LLMs) possess genuine structural reasoning, or merely rely on surface-level pattern matching? The financial domain, demanding numerical precision and mult…

cs.AI2026

Science Edge Evaluation: SEE the Missing Step Toward Real Scientific Discovery

Taolin Han, Yuchen Zhang, Jinghang Wang +22

Large language models (LLMs) are increasingly involved in scientific discovery, yet it remains unclear whether they can support complex real laboratory science. Here we introduce S…

cs.IR2026

UniRank: End-to-End Domain-Specific Reranking of Hybrid Text-Image Candidates

Yupei Yang, Lin Yang, Wanxi Deng +3

Reranking is a critical component in many information retrieval pipelines. Despite remarkable progress in text-only settings, multimodal reranking remains challenging, particularly…

cs.AI2026

Logics-Parsing-Omni Technical Report

Xin An, Jingyi Cai, Xiangyang Chen +22

Addressing the challenges of fragmented task definitions and the heterogeneity of unstructured data in multimodal parsing, this paper proposes the Omni Parsing framework. This fram…

cs.CV2026

Rethinking Cross-Layer Information Routing in Diffusion Transformers

Chao Xu, Maohua Li, Qirui Li +9

Diffusion Transformers (DiTs) have become a de facto backbone of modern visual generation, and nearly every major axis of their design -- tokenization, attention, conditioning, obj…

cs.DC2026

RollArt: Disaggregated Multi-Task Agentic RL Training at Scale

Wei Gao, Yuheng Zhao, Tianyuan Wu +15

Agentic Reinforcement Learning (RL) trains LLMs through multi-turn interactions with environments, producing workloads that mix compute-bound prefill, bandwidth-bound decoding, CPU…

cs.CV2026

FilmBench: A Film-Grade Benchmark for Cinematic Video Generation

Shengyi Wang, Niantong Li, Guangzheng Hu +27

Progress in video generation keeps narrowing the visual gap between AI-generated and professionally produced footage, yet most benchmarks still draw prompts from web sources or LLM…

cs.DC2022

PICASSO: Unleashing the Potential of GPU-centric Training for Wide-and-deep Recommender Systems

Yuanxing Zhang, Langshi Chen, Siran Yang +12

The development of personalized recommendation has significantly improved the accuracy of information matching and the revenue of e-commerce platforms. Recently, it has 2 trends: 1…

cs.CL2026

Hallucination-Resistant Relation Extraction via Dependency-Aware Sentence Simplification and Two-tiered Hierarchical Refinement

Yupei Yang, Fan Feng, Lin Yang +5

Relation extraction (RE) enables the construction of structured knowledge for many downstream applications. While large language models (LLMs) have shown great promise in this task…

cs.AI2026

SPM-Bench: Benchmarking Large Language Models for Scanning Probe Microscopy

Peiyao Xiao, Xiaogang Li, Xinyi Gao +7

As LLMs achieved breakthroughs in general reasoning, their proficiency in specialized scientific domains reveals pronounced gaps in existing benchmarks due to data contamination, i…

cs.CV2026

Qwen-Image-VAE-2.0 Technical Report

Zekai Zhang, Deqing Li, Kuan Cao +27

We present Qwen-Image-VAE-2.0, a suite of high-compression Variational Autoencoders (VAEs) that achieve significant advances in both reconstruction fidelity and diffusability. To a…

cs.IR2026

Prompt Generation Technical Report

Dan Ou, Gui Ling, Hao Wan +25

The paper introduces Prompt Generation (PG), a configuration‑driven framework that separates feature processing from model architecture for generative retrieval systems, enabling f…

#generative retrieval#prompt generation#feature engineering#configuration-driven framework
cs.AI2026

Logics-STEM: Empowering LLM Reasoning via Failure-Driven Post-Training and Document Knowledge Enhancement

Mingyu Xu, Cheng Fang, Keyue Jiang +16

We present Logics-STEM, a state-of-the-art reasoning model fine-tuned on Logics-STEM-SFT-Dataset, a high-quality and diverse dataset at 10M scale that represents one of the largest…

cs.CR2025

MapComp: A Secure View-based Collaborative Analytics Framework for Join-Group-Aggregation

Xinyu Peng, Feng Han, Li Peng +8

Join-group-aggregation (JGA) queries are fundamental to data analytics, yet executing them collaboratively across different parties poses significant privacy risks. Secure multi-pa…

cs.LG2026

Factored Causal Representation Learning for Robust Reward Modeling in RLHF

Yupei Yang, Lin Yang, Wanxi Deng +5

A reliable reward model is essential for aligning large language models with human preferences through reinforcement learning from human feedback. However, standard reward models a…

cs.DB2025

Dynamically Detect and Fix Hardness for Efficient Approximate Nearest Neighbor Search

Zhiyuan Hua, Qiji Mo, Zebin Yao +8

Approximate Nearest Neighbor Search (ANNS) has become a fundamental component in many real-world applications. Among various ANNS algorithms, graph-based methods are state-of-the-a…

cs.DC2026

CoCoScale: Leveraging Layer-wise Scaling to Unlock the Potential of Online LLM Serving

Jingfeng Wu, Yiyuan He, Minxian Xu +7

Online large language model (LLM) serving has become the backbone of modern AI applications, powering diverse downstream services through shared hardware clusters. However, modern…

cs.CV2026

DMC-CF: Dynamic Multimodal CounterFactual QA benchmark for Causal Reasoning

Junzhe Zhang, Huixuan Zhang, Guirong Wang +5

With the rapid advancement of multimodal large language models (MLLMs), models have demonstrated increasingly powerful multimodal capabilities. However, whether MLLMs trained throu…

cs.CL2026

CKV: Compressed and Composable KV Cache Reuse for Efficient LLM Inference

Chuheng Du, Junyi Chen, Hanlin Tang +7

Long-context inference is central to modern large language model (LLM) applications such as retrieval-augmented generation and multi-document reasoning. To mitigate the growing inf…

cs.LG2026

Discovering Millions of Interpretable Features with Sparse Autoencoders

XinYang He, Wei Wang, Bing Zhao +5

Sparse autoencoders (SAEs) have emerged as a powerful tool for decomposing superposed language model representations into sparse and interpretable features. However, training SAEs…

cs.LG2026

Shiva-DiT: Residual-Based Differentiable Top- Selection for Efficient Diffusion Transformers

Jiaji Zhang, Hailiang Zhao, Guoxuan Zhu +9

Diffusion Transformers (DiTs) incur prohibitive computational costs due to the quadratic scaling of self-attention. Existing pruning methods fail to simultaneously satisfy differen…

cs.CV2025

Logics-Parsing Technical Report

Xiangyang Chen, Shuzhao Li, Xiuwen Zhu +7

Recent advances in Large Vision-Language models (LVLM) have spurred significant progress in document parsing task. Compared to traditional pipeline-based methods, end-to-end paradi…

cs.CV2025

Qwen-Image Technical Report

Chenfei Wu, Jiahao Li, Jingren Zhou +36

We present Qwen-Image, an image generation foundation model in the Qwen series that achieves significant advances in complex text rendering and precise image editing. To address th…

cs.DC2026

Spotlight: Synergizing Seed Exploration and Spot GPUs for DiT RL Post-Training

Ruiqi Lai, Dakai An, Wei Gao +6

Reinforcement learning (RL) post-training of Diffusion Transformers (DiTs) is prohibitively expensive, requiring thousands of high-end GPUs. Existing works explore two directions t…

cs.IR2025

RecIS: Sparse to Dense, A Unified Training Framework for Recommendation Models

Hua Zong, Qingtao Zeng, Zhengxiong Zhou +31

In this paper, we propose RecIS, a unified Sparse-Dense training framework designed to achieve two primary goals: 1. Unified Framework To create a Unified sparse-dense training fra…

cs.DC2025

RollPacker: Mitigating Long-Tail Rollouts for Fast, Synchronous RL Post-Training

Wei Gao, Yuheng Zhao, Dakai An +11

Reinforcement Learning (RL) is a pivotal post-training technique for enhancing the reasoning capabilities of Large Language Models (LLMs). However, synchronous RL post-training oft…

cs.LG2025

Reinforcement Learning Optimization for Large-Scale Learning: An Efficient and User-Friendly Scaling Library

Weixun Wang, Shaopan Xiong, Gengru Chen +38

We introduce ROLL, an efficient, scalable, and user-friendly library designed for Reinforcement Learning Optimization for Large-scale Learning. ROLL caters to three primary user gr…

cs.IR2026

UserGPT Technical Report

Yunyi Xuan, Hao Yi, Fengling Mao +9

Personalized user understanding from large-scale digital traces remains a fundamental challenge. Traditional user profiling methods rely on discriminative models and manual feature…

cs.DC2026

ROSE: Rollout On Serving GPUs via Cooperative Elasticity for Agentic RL

Wei Gao, Yuheng Zhao, Dilxat Muhtar +13

Agentic reinforcement learning (RL) is reshaping LLM post-training, but end-to-end training time is dominated by compute-intensive, multi-turn rollouts whose resource demand varies…

cs.CL2026

Full Attention Strikes Back: Transferring Full Attention into Sparse within Hundred Training Steps

Yanke Zhou, Yiduo Li, Hanlin Tang +6

Long-context inference in large language models is bottlenecked by the quadratic cost of full attention. Existing efficient alternatives often rely either on native sparse training…

cs.GT2025

Reliable and Private Utility Signaling for Data Markets

Li Peng, Jiayao Zhang, Yihang Wu +6

The explosive growth of data has highlighted its critical role in driving economic growth through data marketplaces, which enable extensive data sharing and access to high-quality…

cs.DC2025

BanaServe: Unified KV Cache and Dynamic Module Migration for Balancing Disaggregated LLM Serving in AI Infrastructure

Yiyuan He, Minxian Xu, Jingfeng Wu +7

Large language models (LLMs) are increasingly deployed in AI infrastructure, driving the need for high throughput, resource efficient serving systems. Disaggregated LLM serving, wh…

cs.DC2025

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation

Tianyuan Wu, Lunxi Cao, Hanfeng Lu +8

Training large Deep Neural Network (DNN) models at scale often encounters straggler issues, mostly in communications due to network congestion, RNIC/switch defects, or topological…

cs.DC2025

InstGenIE: Generative Image Editing Made Efficient with Mask-aware Caching and Scheduling

Xiaoxiao Jiang, Suyi Li, Lingyun Yang +12

Generative image editing using diffusion models has become a prevalent application in today's AI cloud services. In production environments, image editing typically involves a mask…

cs.DC2024

FALCON: Pinpointing and Mitigating Stragglers for Large-Scale Hybrid-Parallel Training

Tianyuan Wu, Wei Wang, Yinghao Yu +7

Fail-slows, or stragglers, are common but largely unheeded problems in large-scale hybrid-parallel training that spans thousands of GPU servers and runs for weeks to months. Yet, t…

cs.SE2026

Socratic-SWE: Self-Evolving Coding Agents via Trace-Derived Agent Skills

Chuan Xiao, Zhengbo Jiao, Shaobo Wang +5

LLM-driven software engineering agents have become a central testbed for real-world language-model capability, yet their training remains limited by the availability of high-qualit…

cs.CL2026

DLawBench: Evaluating LLMs Through Multi-Turn Legal Consultation

Li Zhang, Yuzhen Shi, Yiran Hu +15

Lawyer-client consultation is a critical starting point for legal services. Effective legal assistance hinges on eliciting sufficient and truthful information from clients in order…

cs.DC2026

LegoDiffusion: Micro-Serving Text-to-Image Diffusion Workflows

Lingyun Yang, Suyi Li, Tianyu Feng +10

Text-to-image generation executes a diffusion workflow comprising multiple models centered on a base diffusion model. Existing serving systems treat each workflow as an opaque mono…

cs.IR2026

Rethinking Retrieval-Augmentation as Synthesis: A Query-Aware Context Merging Approach

Jiarui Guo, Yuemeng Xu, Zongwei Lv +6

Retrieval-Augmented Generation (RAG) enables Large Language Models (LLMs) to extend their existing knowledge by dynamically incorporating external information. However, practical d…

cs.CR2025

Efficient and High-Accuracy Secure Two-Party Protocols for a Class of Functions with Real-number Inputs

Hao Guo, Zhaoqian Liu, Liqiang Peng +4

In two-party secret sharing scheme, values are typically encoded as unsigned integers , whereas real-world applications often require computations on signed real…

cs.CL2024

DDK: Distilling Domain Knowledge for Efficient Large Language Models

Jiaheng Liu, Chenchen Zhang, Jinyang Guo +13

Despite the advanced intelligence abilities of large language models (LLMs) in various applications, they still face significant computational and storage demands. Knowledge Distil…

cs.OS2026

RTP-LLM: High-Performance Alibaba LLM Inference Engine

Boyu Tan, Jiarui Guo, Zongwei Lv +26

Large Language Models (LLMs) have revolutionized AI applications, but deploying them at scale presents significant challenges. We present RTP-LLM, a high-performance inference engi…

cs.CV2026

Qwen-Image-2.0 Technical Report

Bing Zhao, Chenfei Wu, Deqing Li +72

We present Qwen-Image-2.0, an omni-capable image generation foundation model that unifies high-fidelity generation and precise image editing within a single framework. Despite rece…

cs.DC2024

SwiftDiffusion: Efficient Diffusion Model Serving with Add-on Modules

Suyi Li, Lingyun Yang, Xiaoxiao Jiang +12

Text-to-image (T2I) generation using diffusion models has become a blockbuster service in today's AI cloud. A production T2I service typically involves a serving workflow where a b…

cs.CL2026

ClinConsensus: A Physician-Calibrated Benchmark for Evaluating Clinical Rubric Coverage in Chinese Medical LLMs

Xiang Zheng, Han Li, Wenjie Luo +15

Open-ended medical LLM evaluation remains weakly grounded in physician-calibrated coverage of clinically relevant response criteria, especially in localized clinical settings. We i…

cs.LG2026

RT-Lynx: Putting the GEMM Sparsity In a Right Way for Diffusion Models

Xing Cong, Hanlin Tang, Kan Liu +3

Diffusion Transformers (DiT) achieve strong performance in image generation but incur substantial inference costs. While prior work has reduced this cost via quantization and disti…

cs.LG2022

GBA: A Tuning-free Approach to Switch between Synchronous and Asynchronous Training for Recommendation Model

Wenbo Su, Yuanxing Zhang, Yufeng Cai +9

High-concurrency asynchronous training upon parameter server (PS) architecture and high-performance synchronous training upon all-reduce (AR) architecture are the most commonly dep…

cs.CV2026

Text Template Tokens Are Implicit Semantic Registers in Diffusion Transformers

Maohua Li, Qirui Li, Yanke Zhou +10

Modern text-to-image diffusion transformers (DiTs) generate images through joint attention, in which text and image tokens interact directly within a single sequence. In large-scal…

cs.DB2025

TRIM: Accelerating High-Dimensional Vector Similarity Search with Enhanced Triangle-Inequality-Based Pruning

Yitong Song, Pengcheng Zhang, Chao Gao +4

High-dimensional vector similarity search (HVSS) is critical for many data processing and AI applications. However, traditional HVSS methods often require extensive data access for…

cs.DC2025

RollMux: Phase-Level Multiplexing for Disaggregated RL Post-Training

Tianyuan Wu, Lunxi Cao, Yining Wei +11

Rollout-training disaggregation is emerging as the standard architecture for Reinforcement Learning (RL) post-training, where memory-bound rollout and compute-bound training are ph…