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

Atlas / People / Detail

Haibin Lin

Haibin Lin works on LLM infrastructure at ByteDance Seed, focusing on training systems for large language and multimodal models from pre-training to post-training.

LLM infrastructure for large-scale language and multimodal models at ByteDance Seed.1 organizations2 reports

Profile status: updated

updated 3 public sources
LLM InfrastructureDistributed TrainingMultimodal ModelsMachine Learning SystemsNatural Language ProcessingDatabase Systems

Current frame

LLM infrastructure for large-scale language and multimodal models at ByteDance Seed.

Extended note

Haibin Lin is a ByteDance Seed researcher focused on LLM infrastructure and large-scale training systems. His public homepage describes work on optimizing training frameworks for LLMs and multimodal models, spanning large-scale pre-training and post-training reinforcement learning infrastructure. Public profiles and project links connect him with systems and infrastructure work around MegaScale, verl, BytePS, GluonNLP, and Apache MXNet, and his Google Scholar profile lists interests in database systems, machine learning systems, and natural language processing.

Public links

website Personal homepage github GitHub google_scholar Google Scholar x X

Organizations

core ByteDance Seed

Reports

Vision-Language Models Seed1.5-VL Technical Report Report Seed1.5-Thinking: Advancing Superb Reasoning Models with Reinforcement Learning

Official and primary sources

on the Path to AGI Official source · homepage · Google Sites Haibin Lin - Google Scholar Official source · scholar · Google Scholar

Supporting sources

eric-haibin-lin (haibin) · GitHub Supporting source · github · GitHub