Yizhe Zhang
Research scientist at Apple MLR; previously at Meta AI and Microsoft Research; works on LLM training, agents, and dialogue.
- Role
- researcher
- Currently at
- Apple
- twitter.com/YizheZhangNLP
- GitHub
- github.com/dreasysnail
- Scholar
- scholar.google.com/citations
- Papers
- 20
Cite
Notes
Only stored in your browser.
Authored papers
20Embarrassingly Simple Self-Distillation Improves Code Generation
arXiv 2026
Reversal Blessing: Thinking Backward May Outpace Thinking Forward in Multi-choice Questions
arXiv 2025
Learn to Reason Efficiently with Adaptive Length-based Reward Shaping
arXiv 2025
Eliciting In-context Retrieval and Reasoning for Long-context Large Language Models
arXiv 2025
DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Generation
arXiv 2025
CLaRa: Bridging Retrieval and Generation with Continuous Latent Reasoning
arXiv 2025
SAGE: Steering and Refining Dialog Generation with State-Action Augmentation
arXiv 2025
SWE-Gym: An Open Environment for Training Software Engineering Agents and Verifiers
preprint
OpenHands: An Open Platform for AI Software Developers as Generalist Agents
arXiv 2024
Scaling Diffusion Language Models via Adaptation from Autoregressive Models
arXiv 2024
ToolSandbox: A Stateful, Conversational, Interactive Evaluation Benchmark for LLM Tool Use Capabilities
arXiv 2024
How Far Are We from Intelligent Visual Deductive Reasoning?
arXiv 2024
Divide-or-Conquer? Which Part Should You Distill Your LLM?
arXiv 2024
Matryoshka Diffusion Models
arXiv 2023
Probing the Multi-turn Planning Capabilities of LLMs via 20 Question Games
arXiv 2023
Stabilizing Transformer Training by Preventing Attention Entropy Collapse
arXiv 2023
PLANNER: Generating Diversified Paragraph via Latent Language Diffusion Model
planner-generating-diversified-paragraph-via
RetGen: A Joint framework for Retrieval and Grounded Text Generation Modeling
arXiv 2021
POINTER: Constrained Progressive Text Generation via Insertion-based Generative Pre-training
EMNLP 2020 11
DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation
arXiv 2019
Affiliations
Frequent co-authors
10from 20 papers