LLMpeople
Public Atlas People first, reports as evidence, organizations as context.

Atlas / People / Detail

Defa Zhu

Defa Zhu is an AI researcher whose public homepage says his work focuses on large language models and stronger model architectures.

AI researcher at ByteDance focused on large language model architectures.1 organizations1 reports

Profile status: updated

updated 2 public sources
large language models

Current frame

AI researcher at ByteDance focused on large language model architectures.

Extended note

Defa Zhu is an AI researcher at ByteDance whose official homepage says his work focuses on large language models and stronger LLM architectures. OpenReview lists his education history as an MS student at the University of Chinese Academy of Sciences from 2017 to 2020 and an undergraduate student in mathematics at Northeastern University from 2013 to 2017. His homepage highlights work including Hyper-Connections, Frac-Connections, Over-Tokenized Transformer, and Expert Race.

Education

University of Chinese Academy of Sciences MS 2017 → 2020
Northeastern University Mathematics 2013 → 2017

Public links

website Homepage google_scholar Google Scholar openreview OpenReview

Organizations

core ByteDance Seed

Reports

Report Seed1.5-Thinking: Advancing Superb Reasoning Models with Reinforcement Learning

Official and primary sources

Defa Zhu (朱德发) | AI Researcher Official source · homepage Defa Zhu Official source · openreview · OpenReview