TreeSynth: Synthesizing Diverse Data from Scratch via Tree-Guided Subspace Partitioning

Jan 1, 2025·

Sheng Wang

Equal contribution

Pengan Chen

Equal contribution

Jingqi Zhou

Equal contribution

Qintong Li

Jingwei Dong

Jiahui Gao

Boyang Xue

Jiyue Jiang

Lingpeng Kong

Chuan Wu

· 0 min read

Preprint Preprint

Abstract

TreeSynth introduces a novel approach to synthesizing diverse data from scratch via tree-guided subspace partitioning, enabling more effective data generation for machine learning applications.

Type

Conference paper

Publication

In Neural Information Processing Systems (NeurIPS) 2025

Last updated on Jan 1, 2025

Data Synthesis NeurIPS 2025 Spotlight First Author

Authors

Sheng Wang (Forence)

PhD Graduate in Computer Science

Sheng Wang is a PhD graduate from The University of Hong Kong, supervised by Prof. Chuan Wu and Prof. Lingpeng Kong. His research focuses on Agent, LLM Super-Alignment, and Data Synthesis. He has published 14+ papers in top-tier conferences including NIPS2025 (Spotlight), ICLR2025, ACL2024/2025, EMNLP2025.

Authors

Authors

Authors

Authors

Authors

Authors