Zijian Zhou is a former Research Engineer at the Singapore-MIT Alliance for Research and Technology Centre (SMART), M3S program.
His research focuses on data valuation, large language model (LLM) inference-time techniques, and LLM post-training.
Zijian Zhou currently works as an LLM Engineer at MiniMax AI.
He was also awarded the Best Paper Award at the NeurIPS 2025
MineDraft: A Framework for Batch Parallel Speculative Decoding
Year of Publication:
Venue:
ICML 2026
2026
Publication Type:
Conference
Status:
Accepted
Authors:
Zhenwei Tang (co-first authors) (NTU), Arun Verma (co-first authors) (SMART), Zijian Zhou (SMART), Zhaoxuan Wu (SMART), Alok Prakash (SMART), Daniela Rus (SMART & MIT), Bryan Kian Hsiang Low (SMART & NUS)
Year of Publication:
Venue:
ICLR 2026 (Best Paper Award at the Workshop on Multi-Turn Interactions in Large Language Models (MTI-LLM), in conjunction with NeurIPS 2025)
2025
Publication Type:
Conference
Status:
Published
Authors:
Zijian Zhou (SMART), Ao Qu (MIT), Zhaoxuan Wu (SMART), Sunghwan Kim (Yonsei University), Alok Prakash (SMART), Daniela Rus (MIT), Jinhua Zhao (MIT), Bryan Kian Hsiang Low (NUS), Paul Pu Liang (MIT)
Year of Publication:
Venue:
The 63rd Annual Meeting of the Association for Computational Linguistics (ACL), Vienna, Austria, 27th July, 2025
2025
Publication Type:
Conference
Status:
Published
Authors:
Zhaoxuan Wu (SMART), Zijian Zhou (SMART), Arun Verma (SMART), Alok Prakash (SMART), Daniela Rus (MIT), and Bryan Kian Hsiang Low (NUS)
Year of Publication:
Venue:
Annual Conference on Neural Information Processing Systems (NeurIPS)
2024
Publication Type:
Conference
Status:
Published
Authors:
Zijian Zhou (SMART), Xiaoqiang Lin (NUS), Xinyi Xu (NUS), Alok Prakash (SMART), Daniela Rus (MIT), Bryan Low (NUS)