XG
Xiaodong Gu
cs.SEcs.CLcs.AIcs.LGcs.GRcs.PLcs.NEcs.CVcs.CRcs.IR
On Valency
published · living versionsW_fppdawyn·v1 · currentpublished
Reasoning Patterns for Efficient AI Coding: An Empirical Study
with Zhe Yin, Yuling Shi, Beijun Shen
1 version
Preprints & journals
78 papers in the corpus · 2016–2026SWE-at-Scale: Scalable Issue Synthesis from Historical Bug Fixes2610.21359v1 · Silin Chen, Zihan Zhang, Xiaodong Gu et al.2026 · 0 citationsarXiv
Translator vs. Challenger: Adversarial Agentic Learning for C-to-Rust Translation2609.15381v1 · Chaofan Wang, Xiaodong Gu, Yuling Shi et al.2026 · 0 citationsarXiv
EarlyEval: Cheaper Agent Evaluation via Early Outcome Prediction2609.02783v1 · Yuling Shi, Zhensu Sun, Junsen Dong et al.2026 · 0 citationsarXiv
DeepRepoQA: Code Repository Question Answering with Deep Agent Exploration2608.24221v1 · Weihan Peng, Yuling Shi, Yingwei Ma et al.2026 · 0 citationsarXiv
SRPO: Self-Reflective Policy Optimization for Long-Horizon Reasoning2608.23493v1 · Jialong Liu, Yuling Shi, Ning Yang et al.2026 · 0 citationsarXiv
Schrödinger's Code Repository: Have LLMs Learned SWE-bench or Memorized It?2609.27891v1 · Silin Chen, Yufei Yang, Xiaodong Gu et al.2026 · 0 citationsarXiv
Repo0: Design-Driven Zero-to-All Code Generation2608.19854v1 · Silin Chen, Haoyi Teng, Xiaodong Gu et al.2026 · 0 citationsarXiv
SkillForge: Self-Distilling Agents for Project-Specific Issue Resolution2608.18933v1 · Silin Chen, Han Li, Xiaodong Gu et al.2026 · 0 citationsarXiv
ParaTempo: Efficient Parallel Reasoning via Temporal Confidence2608.16425v1 · Xuteng Zhang, Wenhao Zeng, Xiaodong Gu et al.2026 · 0 citationsarXiv
SWE-Pruner Pro: The Coder LLM Already Knows What to Prune2607.18213v1 · Yuhang Wang, Yuling Shi, Shaoqiu Zhang et al.2026 · 0 citationsarXiv
Know Before Fix: QA-Driven Repository Knowledge Acquisition for Software Issue Resolution2607.11111v1 · Haotian Lin, Silin Chen, Xiaodong Gu et al.2026 · 0 citationsarXiv
Dockerless: Environment-Free Program Verifier for Coding Agents2606.28436v1 · Wenhao Zeng, Yuling Shi, Xiaodong Gu et al.2026 · 0 citationsarXiv
Automated jailbreak attack targeting multiple defense strategies2606.16751v1 · Qi Wang, Chengcheng Wan, Weijia He et al.2026 · 0 citationsarXiv
LLM Agents Can See Code Repositories2606.14061v4 · Dongjian Ma, Silin Chen, Yufei Yang et al.2026 · 0 citationsarXiv
FastContext: Training Efficient Repository Explorer for Coding Agents2606.14066v4 · Shaoqiu Zhang, Maoquan Wang, Yuling Shi et al.2026 · 0 citationsarXiv
Code Is More Than Text: Uncertainty Estimation for Code Generation2606.09577v1 · Yuling Shi, Caiqi Zhang, Yuexian Li et al.2026 · 0 citationsarXiv
SWE-Explore: Benchmarking How Coding Agents Explore Repositories2606.07297v1 · Shaoqiu Zhang, Yuhang Wang, Jialiang Liang et al.2026 · 2 citationsarXiv
Reward-Decomposed Reinforcement Learning for Immersive Video Role-Playing2605.04733v3 · Miao Wang, Yuling Shi, Yijiang Li et al.2026 · 0 citationsarXiv
ClassEval-Pro: A Cross-Domain Benchmark for Class-Level Code Generation2604.26923v1 · Yeheng Chen, Chaoxiang Xie, Yuling Shi et al.2026 · 0 citationsarXiv
EffiSkill: Agent Skill Based Automated Code Efficiency Optimization2603.27850v1 · Zimu Wang, Yuling Shi, Mengfan Li et al.2026 · 1 citationarXiv
Rethinking Code Complexity Through the Lens of Large Language Models2602.07882v2 · Chen Xie, Xiaodong Gu, Yuling Shi et al.2026 · 0 citationsarXiv
Rethinking the Value of Agent-Generated Tests for LLM-Based Software Engineering Agents2602.07900v2 · Zhi Chen, Zhensu Sun, Yuling Shi et al.2026 · 0 citationsarXiv
Test vs Mutant: Adversarial LLM Agents for Robust Unit Test Generation2602.08146v3 · Pengyu Chang, Yixiong Fang, Silin Chen et al.2026 · 0 citationsarXiv
Seeing is Coding: On the Effectiveness of Vision Language Models in Code Understanding2602.01785v3 · Yuling Shi, Chaoxiang Xie, Zhensu Sun et al.2026 · 1 citationarXiv
SWE-Pruner: Self-Adaptive Context Pruning for Coding Agents2601.16746v4 · Yuhang Wang, Yuling Shi, Mo Yang et al.2026 · 2 citationsarXiv
Reasoning in Trees: Improving Retrieval-Augmented Generation for Multi-Hop Question Answering2601.11255v1 · Yuling Shi, Maolin Sun, Zijun Liu et al.2026 · 2 citationsarXiv
Readability-Robust Code Summarization via Meta Curriculum Learning2601.05485v2 · Wenhao Zeng, Yitian Chai, Hao Zhou et al.2026 · 0 citationsarXiv
GlimpRouter: Efficient Collaborative Inference by Glimpsing One Token of Thoughts2601.05110v3 · Wenhao Zeng, Xuteng Zhang, Yuling Shi et al.2026 · 0 citationsarXiv
CatchAll: Repository-Aware Exception Handling with Knowledge-Guided LLMs2601.01271v1 · Qingxiao Tao, Xiaodong Gu, Hao Zhong et al.2026 · 0 citationsarXiv
In Line with Context: Repository-Level Code Generation via Context Inlining2601.00376v3 · Chao Hu, Wenhao Zeng, Yuling Shi et al.2026 · 1 citationarXiv
Analyzing the Mechanism of Attention Collapse in VGGT from a Dynamics Perspective2512.21691v1 · Huan Li, Longjun Luo, Yuling Shi et al.2025 · 0 citationsarXiv
Neuron-Guided Interpretation of Code LLMs: Where, Why, and How?2512.19980v2 · Zhe Yin, Xiaodong Gu, Beijun Shen2025 · 2 citationsarXiv
Beyond Language Boundaries: Uncovering Programming Language Families for Code Language Models2512.19509v1 · Shangbo Yun, Xiaodong Gu, Jianghong Huang et al.2025 · 0 citationsarXiv
Empowering LLMs with Structural Role Inference for Zero-Shot Graph Learning2511.00898v2 · Heng Zhang, Jing Liu, Jiajun Wu et al.2025 · 0 citationsarXiv
GraphGeo: Multi-Agent Debate Framework for Visual Geo-localization with Heterogeneous Graph Neural Networks2511.00908v2 · Heng Zheng, Yuling Shi, Xiaodong Gu et al.2025 · 0 citationsarXiv
Robust Preference Alignment via Directional Neighborhood Consensus2510.20498v3 · Ruochen Mao, Yuling Shi, Xiaodong Gu et al.2025 · 0 citationsarXiv
GraphShaper: Geometry-aware Alignment for Improving Transfer Learning in Text-Attributed Graphs2510.12085v2 · Heng Zhang, Tianyi Zhang, Yuling Shi et al.2025 · 0 citationsarXiv
Can Representation Gaps Be the Key to Enhancing Robustness in Graph-Text Alignment?2510.12087v1 · Heng Zhang, Tianyi Zhang, Yuling Shi et al.2025 · 0 citationsarXiv
GraphTracer: Graph-Guided Failure Tracing in LLM Agents for Robust Multi-Turn Deep Search2510.10581v2 · Heng Zhang, Yuling Shi, Xiaodong Gu et al.2025 · 0 citationsarXiv
D3MAS: Decompose, Deduce, and Distribute for Enhanced Knowledge Sharing in Multi-Agent Systems2510.10585v3 · Heng Zhang, Yuling Shi, Xiaodong Gu et al.2025 · 0 citationsarXiv
HyperAgent: Leveraging Hypergraphs for Topology Optimization in Multi-Agent Communication2510.10611v3 · Heng Zhang, Yuling Shi, Xiaodong Gu et al.2025 · 0 citationsarXiv
LongCodeZip: Compress Long Context for Code Language Models2510.00446v1 · Yuling Shi, Yichun Qian, Hongyu Zhang et al.2025 · 6 citationsarXiv
Attention as a Compass: Efficient Exploration for Process-Supervised RL in Reasoning Models2509.26628v1 · Runze Liu, Jiakang Wang, Yuling Shi et al.2025 · 0 citationsarXiv
SWE-QA: Can Language Models Answer Repository-level Code Questions?2509.14635v2 · Weihan Peng, Yuling Shi, Yuhang Wang et al.2025 · 2 citationsarXiv
ALT4Decompile: Inferring C-aligned Abstract Loop Tree for LLM-Based Binary Decompilation2509.14646v2 · Yongpan Wang, Puzhuo Liu, Xin Xu et al.2025 · 1 citationarXiv
Transplant Then Regenerate: A New Paradigm for Text Data Augmentation2508.14723v3 · Guangzhan Wang, Hongyu Zhang, Beijun Shen et al.2025 · 0 citationsarXiv
LibRec: Benchmarking Retrieval-Augmented LLMs for Library Migration Recommendations2508.09791v1 · Junxiao Han, Yarong Wang, Xiaodong Gu et al.2025 · 1 citationarXiv
Pruning the Unsurprising: Efficient LLM Reasoning via First-Token Surprisal2508.05988v2 · Wenhao Zeng, Yaoning Wang, Chao Hu et al.2025 · 1 citationarXiv
EvoC2Rust: A Skeleton-guided Framework for Project-Level C-to-Rust Translation2508.04295v4 · Chaofan Wang, Tingrui Yu, Beijun Shen et al.2025 · 5 citationsarXiv
SWE-Debate: Competitive Multi-Agent Debate for Software Issue Resolution2507.23348v1 · Han Li, Yuling Shi, Shaoxin Lin et al.2025 · 3 citationsarXiv
Career total: 122 works. 78 are in this corpus.Showing the 50 most recent.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.