Biography
Greetings! I’m Yifan Zhang (中文: 张一凡, 日本語: チャン・イーファン, Русский: Ифань Чжан), a Senior Research Scientist at Salesforce AI Research (SFR) in San Francisco/Palo Alto. I received my Ph.D. in Computer Science from Vanderbilt University (Vandy), where I was supervised by Prof. Kevin J. Leach and Prof. Yu Huang. I also received my M.Sc. in Computer Science from Georgia Institute of Technology (GaTech). Previously, I earned my B.A., B.Eng., and M.Eng. from China University of Petroleum (CUP) in Beijing and worked as a full-time MLE at JD.com. During my doctoral studies, I completed research internships at Google Research, Amazon AWS, IBM Research, Intel Corporation, and TikTok/ByteDance. My research focuses on leveraging AI for Programming Languages and Software Engineering (AI4Code), with a particular emphasis on combining program analysis and symbolic reasoning with interactive AI systems to improve neural code analysis, comprehension, and software engineering automation.
- Interactive and Agentic AI
- Neuro-Symbolic AI for PL/SE
- Data-Centric AI for Systems
- LLVM IR and Binary Code Analysis
- Compiler Optimization
- Applied NLP/ML
Vanderbilt University, 2022-2026
Ph.D. in Computer Science, AI for Software Engineering
Georgia Institute of Technology, 2022-2026
M.Sc. in Computer Science, Artificial Intelligence
China University of Petroleum, 2012-2019
B.A., B.Eng., M.Eng. in English & Petroleum Engineering
University of Calgary, 2015-2016
Undergraduate Exchange Program in Petroleum Engineering
News
2026
[09/24/2026] One paper got accepted to NeurIPS'26 with AR of 25.7%.
[08/12/2026] Invited as a reviewer for TSE.
[06/28/2026] One paper got accepted to TOSEM with IF of 6.9.
[06/16/2026] One paper got accepted to AI4Math@ICML'26.
[06/01/2026] Received the Oral Presentation for ACL'26 with OR of 3.9%.
[05/08/2026] Graduated from Vanderbilt University with a Ph.D. in Computer Science.
[04/15/2026] Invited as a reviewer for ICSE'27 Demo.
[04/06/2026] One paper got accepted to ACL'26 with AR of 19%.
[04/06/2026] One paper got accepted to ACL'26 Findings.
[04/05/2026] Received the Great Review Distinction for ACL ARR'26 in January.
[03/25/2026] One paper got accepted to FSE'26 with AR of 26.41%.
[03/24/2026] Invited as a reviewer for NeurIPS'26.
[03/11/2026] Passed my Ph.D. Dissertation Defense.
[01/19/2026] Invited as a reviewer for ACL ARR'26 in January.
[01/05/2026] Started my full-time RS at Salesforce AI Research in Palo Alto, CA, USA.
2025
[12/26/2025] One paper got accepted to ICPC'26 with AR of 26.8%.
[11/07/2025] One paper got accepted to AAAI'26 with AR of 17.6%.
[11/03/2025] Invited as a reviewer for LLM4Code@ICSE'26.
[10/27/2025] Attended LLVM Developers’ Meeting in Santa Clara, CA, USA.
[10/24/2025] Passed my Ph.D. Qualifying Examination.
[10/11/2025] One paper got accepted to TMLR.
[09/22/2025] Invited as a reviewer for ICLR'26.
[09/15/2025] Started my AS Intern at Amazon AWS in Arlington, VA, USA.
[07/12/2025] Invited as a reviewer for AAAI'26.
[06/22/2025] Attended FSE'25 in Trondheim, Trøndelag, Norway.
[06/01/2025] Received the NSF Student Travel Support for FSE'25/ISSTA'25.
[05/19/2025] Started my Research Intern at IBM Research in Yorktown Heights, NY, USA.
[05/15/2025] One paper got accepted to ACL'25 with apx. AR of 20%.
[05/06/2025] Received the C.F. Chen Best Paper Award from Vanderbilt University.
[04/27/2025] Attended ICSE'25 in Ottawa, ON, Canada.
[04/09/2025] Received ACM SIGSOFT CAPS Travel Grant for FSE'25/ISSTA'25.
[04/01/2025] Attended HotSoS'25 virtually hosted by NSA.
[03/31/2025] One paper got accepted to EuroS&P'25 with AR of 20%.
[03/13/2025] Invited as a reviewer for ACL'25 Industry.
[03/12/2025] One paper got accepted to FSE'25 IVR.
[02/19/2025] One paper got accepted to HotSoS'25 WIP hosted by NSA.
[02/17/2025] Invited as a reviewer for NeurIPS'25.
[02/13/2025] One paper got accepted to NAACL'25 Industry with apx. AR of 23.2%.
[01/13/2025] Started my Student Researcher at Google LLC in Sunnyvale, CA, USA.
[01/12/2025] One paper got accepted to ICPC'25 ERA.
2024
[12/11/2024] Invited as a reviewer for ICML'25.
[12/10/2024] One paper got accepted to ICSE'25 JF.
[10/15/2024] One paper got accepted to TMech with IF of 7.3.
[08/23/2024] Invited as a reviewer for ICLR'25.
[07/15/2024] Attended FSE'24 in Porto de Galinhas, PE, Brazil.
[06/10/2024] Started my GenAI Intern at Intel Corporation in Santa Clara, CA, USA.
[06/02/2024] Invited as a reviewer for NeurIPS'24 D&B.
[05/30/2024] Invited as a reviewer for DMLR@ICML'24.
[05/24/2024] Invited as a reviewer for NeurIPS'24.
[04/23/2024] One paper got accepted to TOSEM with IF of 6.6.
[04/09/2024] Invited as a reviewer for CCS'24 AE.
[03/15/2024] Selected as a top reviewer for DMLR@ICLR'24 with TR of 9.41%.
[01/25/2024] Invited as a reviewer for MICCAI'24.
[01/23/2024] One paper got accepted to WWW'24 with AR of 20.2%.
[01/22/2024] One paper got accepted to FSE'24 with direct AR of 11.59%.
[01/11/2024] Invited as a reviewer for DMLR@ICLR'24.
[01/04/2024] One paper got accepted to ICPC‘24 RENE.
2023
[12/19/2023] Passed my Ph.D. Preliminary Examination.
[09/05/2023] One paper got accepted to TMI with IF of 10.6.
[07/13/2023] Invited as a reviewer for AAAI'24.
[07/07/2023] One paper got accepted to ASE'23 NIER.
[05/22/2023] Started my RS intern at ByteDance in San Jose, CA, USA.
[03/19/2023] Invited as a reviewer for MLSys'23 AE.
[03/02/2023] One paper got accepted to InteNSE@ICSE'23.
[02/06/2023] Invited as a reviewer for MICCAI'23.
[01/06/2023] Invited as a reviewer for DCAA@AAAI'23.
2022
[12/04/2022] Invited as a shadow reviewer for MSR'23.
[11/28/2022] Attended NeurIPS'22 in New Orleans, LA, USA.
[10/10/2022] Attended ASE'22 in Detroit, MI, USA.
[09/22/2022] Attended MICCAI'22 virtually in Singapore.
[08/05/2022] One paper got accepted to ASE'22 DS.
[08/01/2022] Invited as a reviewer for AAAI'23.
[07/18/2022] One paper got accepted to CaPTion@MICCAI'22.
[07/02/2022] Invited as a reviewer for CaPTion@MICCAI'22.
[06/01/2022] Started my Ph.D. in Computer Science at Vanderbilt University.
Experience
Computer-Use Agents and Coding Agents
- Developed CLIFT, training a Gemma 4 31B web agent with LoRA using conformally certified, URL-aware rewards and reusing the verifier for judge-free test-time selection. Raised WebArena Infinity success from 61.8% to 74.6% and improved Online Mind2Web by 11.3 points on GPT-5.5 (NeurIPS'26).
- Developed DarwinX, evolving agent harnesses around a frozen model through regression-aware selection, independent confirmation, and cross-lineage recombination. Improved Terminal-Bench 2.1 and DeepSWE success by 7.2 and 9.4 percentage points, and reached 91.2% audit-clean pass@1 on WebArena Infinity.
- Contributed to verbal critiques for coding agents (Opera) and training data for harness-model co-evolution (CoTrace).
Automated Reasoning Group, Stateful Cloud API Simulation
- Developed CloudTwin, an LLM-backed AWS API simulator integrated with LocalStack for Terraform testing without live cloud provisioning.
- Designed OASIS, a bounded operation-as-state memory for consistent replay of resource IDs and dependencies across long API traces.
- Achieved 91.4% project-level terraform apply success across 151 projects, versus 36.4% for LocalStack.
AI Reasoning Group, Grounded and Reflective Agent Planning
- Developed SPIRAL, an LLM planning framework that embeds Planner, Simulator, and Critic roles in Monte Carlo Tree Search.
- Used predicted action outcomes and dense reflective feedback to guide multi-step tool-use planning without model fine-tuning.
- Achieved 83.6% cascaded accuracy on TaskBench DailyLifeAPIs with Llama 4 Maverick, versus 67.4% for LATS (AAAI'26).
Google TAO Team, Google CORE Developer
- Designed an automated pipeline with Google CORE/Google Research that uses LLMs to generate C++ code that simulates LLVM IR profiles.
- Enabled benchmarking and performance analysis of production code through the automated pipeline.
- The system supported AI models in learning compiler behavior and optimization strategies across diverse hardware platforms, with safeguards to preserve IP integrity.
Incubation and Disruptive Innovation (IDI) Group, Intel Silicon Valley
- Developed CONSTRUCTA to automate scheduling for semiconductor fabrication facilities, combining construction-specific retrieval, dependency-aware context sampling, and DPO-based alignment with expert preferences.
- Evaluated on 4,340 construction activities, improving automated-planning accuracy by 28.9 percentage points over a GPT-4o basic-context baseline (NAACL'25 Industry).
Infrastructure Systems Lab, TikTok Applied Research Center
- Enhanced ByteBrain’s SQL pipeline with a representation-learning system for recommending optimizer hints and reranking execution plans across diverse table spaces and schemas.
- Designed a two-Transformer workflow and pretrained a generalizable plan reranker for the query-optimization stage downstream of SQL generation.
- Gave a technical talk at TikTok InfraLab and contributed to the team’s LLM literature reviews and conference-paper reviews.
Duke Center for AI in Radiology, Duke University Medical Center
- Developed tumor-insertion augmentation and a lightweight ResNet-ViT detector for breast tomosynthesis, improving lesion-detection AP50 by 13.1% (MICCAIW'22).
- Co-developed SWSSL, a sliding-window self-supervised method for high-resolution anomaly detection that improved image-level AUC on breast tomosynthesis and chest X-rays (IEEE TMI'23).
HKU Business School
- Developed an intelligent debt-collection system that uses Transformer-based representations of customer interaction histories and offline policy-gradient learning to recommend long-horizon collection strategies.
- Modeled evolving account states to support context-aware decisions across multi-step workflows.
Intelligent Risk Management Lab, JD Technology
- Built Bi-GRU and DeepFM models on user behavior features to predict credit use rate and profit, achieving a 21.4% overall profit increase.
- Built a heterogeneous graph and applied GNN models to propagate credit scores, improving XGB user-classification accuracy by 5%.
- Invented a GAN-styled model using differential privacy for federated learning. Granted 5 CN patents, with one awarded 1st Runner-up in the JD Discovery Cup Patent Competition.
Institute for Data Science, Udacity China
- Taught VIP students Python, statistics, machine learning and data mining
- Organized online community activities and created instructional materials
- Awarded Oustanding Mentor of Udacity China (Top 1%)
Funds & Awards
Services
Academic Mentorship
- 2025 Zichen (Ryan) Zhu, undergraduate student in CS & Math & Political Science at Vanderbilt University
- 2024 Suad Mohamed, undergraduate student in CS & Psychology at Belmont University (co-mentored by Mr. Zachary Karas)
- 2024 Manish Acharya, undergraduate student in CS & Math at Vanderbilt University
- 2023 Jieyu Li, bachelor in EEE at Shanghai Jiaotong University, master in ECE at Vanderbilt University
- 2023 Luka Mushkudiani, undergraduate student in CS & Math at Vanderbilt University
- 2022 Eric Li, undergraduate student in CS & Math at Vanderbilt University
Program Committee Member and Conference Reviewer
- 2026 International Workshop on Large Language Models for Code (LLM4Code@ICSE'26)
- 2026 AAAI Conference on Artificial Intelligence (AAAI'26)
- 2025 Annual Meeting of the Association for Computational Linguistics (ACL'25) Industry Track
- 2025 Annual Conference on Neural Information Processing Systems (NeurIPS'25) Datasets & Benchmarks Track
- 2025 International Conference on Machine Learning (ICML'25)
- 2025 International Conference on Learning Representations (ICLR'25)
- 2024 Annual Conference on Neural Information Processing Systems (NeurIPS'24) Datasets & Benchmarks Track
- 2024 Data-centric Machine Learning Research Workshop at ICML (DMLR@ICML'24)
- 2024 Annual Conference on Neural Information Processing Systems (NeurIPS'24)
- 2024 ACM Conference on Computer and Communications Security (CCS'24) Artifact Evaluation Track
- 2024 Data-centric Machine Learning Research Workshop at ICLR (DMLR@ICLR'24)
- 2024 International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI'24)
- 2024 AAAI Conference on Artificial Intelligence (AAAI'24)
- 2023 Annual Conference on Machine Learning and Systems (MLSys'23) Artifact Evaluation Track
- 2023 International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI'23)
- 2023 AAAI Workshop on DL-Hardware Co-Design for AI Acceleration (DCAA@AAAI'23)
- 2023 International Conference on Mining Software Repositories (MSR'23)
- 2023 AAAI Conference on Artificial Intelligence (AAAI'23)
- 2022 MICCAI Workshop on Cancer Prevention through early detecTion (CaPTion@MICCAI'22)
External Conference Reviewer
- 2024 International Conference on Very Large Data Bases (VLDB'24)
- 2023 USENIX Security Symposium (USENIX'23)
- 2022 Computer Vision and Pattern Recognition Conference (CVPR'22)
Contact
- yifan.zhang.2@vanderbilt.edu
- +1 919 638-0501
- 1400 18th Ave S, Nashville, TN 37212
- DM Me on Twitter