Yihong Dong
Assistant Professor
Works on intelligent software engineering with large language models and multi-agent collaboration for code generation.
Yihong Dong's lab at SJTU works on large language models, reasoning & agents and reasoning & neuro-symbolic AI. 70 papers since 2023, 20 at top venues, mostly ACL and Findings of ACL. Applications close .
At a glance
- papers, 2023–2026
- 18/ year
- at T1 venues
- 29%
- PhD students
- —
- MS students
- —
- alumni listed
- —
Papers by yearfrom DBLP, 2023–2026
- 15 papers2026 · T1 7 · collecting
- 28 papers2025 · T1 11
- 14 papers2024 · T1 2
- 13 papers2023 · T1 0
T1 venuesOtherVenues: ACL, Findings of ACL, NeurIPS, Expert Systems with Applications, ICASSP
Research focuspaper titles that use each topic's words
- Large language models20 papers
- code generation17 papers
- Reasoning & agents5 papers
- Reasoning & neuro-symbolic AI4 papers
- software engineering1 paper
Paper list
From DBLP.
202615 papers
- Concurr. Comput. Pract. Exp.Attributed Network Representation Learning Based on Graph Neural Network: A Comprehensive Survey (opens in a new tab)
- ACL (T1)CODERL+: Improving Code Generation via Reinforcement with Execution Semantics Alignment (opens in a new tab)
- ICLR (T1)Detecting Data Contamination from Reinforcement Learning Post-training for Large Language Models (opens in a new tab)
- Findings of ACLEvoCoT: Overcoming the Exploration Bottleneck in Reinforcement Learning for LLMs (opens in a new tab)
- ACM Trans. Softw. Eng. Methodol.Exploring Data-Efficient Adaptation of Large Language Models for Code Generation (opens in a new tab)
- Findings of ACLIntentCoding: Amplifying User Intent in Code Generation (opens in a new tab)
- ACL (T1)KoCo-Bench: Can Large Language Models Leverage Domain Knowledge in Software Development? (opens in a new tab)
- AAAI (T1)Large Language Model Unlearning for Source Code (opens in a new tab)
- IJCAI/ECAIMitigating Collaboration Degeneration in Multi-Agent Code Generation via a Controllable Competitive Collaboration Approach (opens in a new tab)
- Proc. ACM Softw. Eng.RealBench: A Repo-Level Code Generation Benchmark Aligned with Real-World Software Development Practices (opens in a new tab)
- Expert Systems with ApplicationsRegularized spatio-temporal weighted graph convolution network based on multimodal fusion for mild cognitive impairment detection (opens in a new tab)
- ACL (T1)RL-PLUS: Countering Capability Boundary Collapse of LLMs in Reinforcement Learning with Hybrid-policy Optimization (opens in a new tab)
202528 papers
- Concurr. Comput. Pract. Exp.A Novel Adaptive Graph Neural Network Based on Meta-Learning for Cross-Domain Few-Shot Learning (opens in a new tab)
- ICASSP (T1)A Novel Audio-Visual Multimodal Semi-Supervised Model Based on Graph Neural Networks for Depression Detection (opens in a new tab)
- ICICAdaptive Cross-Variable Spectral Filtering for Time Series Forecasting (opens in a new tab)
- ASEAligning LLMs to Fully Utilize the Cross-file Context in Repository-level Code Completion (opens in a new tab)
- MICCAI (T1)Alzheimer's Disease Recognition Based on Adaptive Graph Normalization Flow for Incomplete Multimodal Data Fusion (opens in a new tab)
- Knowl. Based Syst.An Adaptive Dual-channel Multi-modal graph neural network for few-shot learning (opens in a new tab)
- COLING (T1)BANER: Boundary-Aware LLMs for Few-Shot Named Entity Recognition (opens in a new tab)
- ACL (T1)CodeDPO: Aligning Code Models with Self Generated and Verified Source Code (opens in a new tab)
- ACM Trans. Softw. Eng. Methodol.CodeScore: Evaluating Code Generation by Learning Code Execution (opens in a new tab)
- NeurIPS (T1)FAN: Fourier Analysis Networks (opens in a new tab)
- Inf. FusionFC-HGNN: A heterogeneous graph neural network based on brain functional connectivity for mental disorder identification (opens in a new tab)
- Findings of ACLFocused-DPO: Enhancing Code Generation Through Focused Preference Optimization on Error-Prone Points (opens in a new tab)
202414 papers
- Biomed. Signal Process. Control.A detail-oriented super-2D network for pulmonary artery segmentation (opens in a new tab)
- Neural Process. Lett.A Graph Contrastive Learning Model Based on Structural and Semantic View for HIN Recommendation (opens in a new tab)
- PKDDA Novel Population Graph Neural Network Based on Functional Connectivity for Mental Disorders Detection (opens in a new tab)
- Biomed. Signal Process. Control.A novel study for depression detecting using audio signals based on graph neural network (opens in a new tab)
- Comput. Animat. Virtual WorldsAnimation line art colorization based on the optical flow method (opens in a new tab)
- Findings of ACLDevEval: A Manually-Annotated Code Generation Benchmark Aligned with Real-World Code Repositories (opens in a new tab)
- NeurIPS (T1)EvoCodeBench: An Evolving Code Generation Benchmark with Domain-Specific Evaluations (opens in a new tab)
- Findings of ACLGeneralization or Memorization: Data Contamination and Trustworthy Evaluation for Large Language Models (opens in a new tab)
- Findings of ACLPACE: Improving Prompt with Actor-Critic Editing for Large Language Model (opens in a new tab)
- Frontiers Artif. Intell.Re-tear after arthroscopic rotator cuff repair can be predicted using deep learning algorithm (opens in a new tab)
- ACM Trans. Softw. Eng. Methodol.Self-Collaboration Code Generation via ChatGPT (opens in a new tab)
- ACM Trans. Softw. Eng. Methodol.Self-Planning Code Generation with Large Language Models (opens in a new tab)
202313 papers
- Mach. Vis. Appl.A dual-path U-Net for pulmonary vessel segmentation method based on lightweight 3D attention (opens in a new tab)
- Appl. Intell.A gated graph attention network based on dual graph convolution for node embedding (opens in a new tab)
- Biomed. Signal Process. Control.A multi-scale interactive U-Net for pulmonary vessel segmentation method based on transfer learning (opens in a new tab)
- Expert Systems with ApplicationsA time sequence coding based node-structure feature model oriented to node classification (opens in a new tab)
- PRCVAn Audio Correlation-Based Graph Neural Network for Depression Recognition (opens in a new tab)
- ECAIAntecedent Predictions Are More Important Than You Think: An Effective Method for Tree-Based Code Generation (opens in a new tab)
- ISSTACODEP: Grammatical Seq2Seq Model for General-Purpose Code Generation (opens in a new tab)
- Appl. Intell.Fusing heterogeneous information for multi-modal attributed network embedding (opens in a new tab)
- J. Knowl. Manag.Linking online and offline intergenerational knowledge transfer to younger employees' innovative work behaviors: evidence from Chinese hospitals (opens in a new tab)
- Neural Comput. Appl.Multimodal heterogeneous graph attention network (opens in a new tab)
- Appl. Intell.Preference-corrected multimodal graph convolutional recommendation network (opens in a new tab)
- InternetwareSeq2Seq or Seq2Tree: Generating Code Using Both Paradigms via Mutual Learning (opens in a new tab)
Where alumni went
Alumni are not listed on the lab homepage.
Apply to SJTUFall 2027
Computer Science PhD (international applicants)
- Round
- Self-funded
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- USNYU
Samuel Bowman
- Large language models
- USCMU
Daniel Fried
- Large language models
- CAToronto
Ewan Dunbar
- Speech recognition & spoken language
- GBUCL
Pontus Stenetorp
- Generative & foundation models
- USNYU
Chen Zhao
- Reasoning & agents
- CNZhejiang
Haitao Xu
- Large language models
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The member list could not be found on the lab site.
- The alumni list could not be found on the lab site.
- Collected
- Verified
- Papers
- DBLP
- Members
- Not collected
- Alumni
- Not collected
Tiers are venue classifications, not lab rankings.
SJTUcloses