Tianren Ma

马天任

Ph.D. Candidate

Learning and Machine Perception Lab (LAMP)
School of Electronic, Electrical and Communication Engineering
University of Chinese Academy of Sciences (UCAS)
Beijing, China, 100083.

博士研究生

机器学习与感知实验室 (LAMP)
电子电气与通信工程学院
中国科学院大学 (UCAS)
中国，北京，100083

Email: matianren18##mails.ucas.ac.cn
Github: https://github.com/martian422

My Biography

I am a Ph.D. candidate of LAMP at UCAS, advised by Prof. Qixiang Ye.

I got my B.E. degree in UCAS in 2022.

Currently, I'm exploring multimodal model's generation and reinforcement learning methods with discrete structures.

I'm also interested in photography, graphic design, and musicals.

个人简介

我是中国科学院大学LAMP实验室的博士研究生，导师是叶齐祥教授。

2022年，我在中国科学院大学获得工学学士学位。

目前，我正在研究基于离散变量的多模态生成和强化学习办法。

我对摄影，平面设计以及音乐剧都很感兴趣。

Experience

Inclusion AI, Ant Group. Research Intern (AntStar) Mar. 2026 - Present
Working on native multimodal models. Developing unified autoencoders.

Intelligent Creation, ByteDance. Research Intern Aug. 2025 - Feb. 2026
Focus on image editing and assessment. VAE design, VLM post-training and RL.

Jiutian Research, CMRI. Research Intern Mar. 2024 - Jul. 2024
Focus on referential comprehension with MLLMs.

Undergraduate Developer Sep. 2021 - Jun. 2022
Developed few-shot algorithms for remote sensing satellites, currently operating in-orbit.

经历

蚂蚁集团 Inclusion AI 研究实习生（AntStar 项目） 2026年3月 - 至今
从事原生多模态模型研究，开发统一自编码器。

字节跳动 Intelligent Creation 研究实习生 2025年8月 - 2026年2月
研究图像编辑与评估，包括VAE设计、VLM后训练与强化学习。

九天人工智能研究院 研究实习生 2024年3月 - 2024年7月
研究多模态大模型中的指代理解问题。

本科开发者 2021年9月 - 2022年6月
开发遥感卫星小样本算法，目前该算法已在轨运行。

Major Publications

主要论文

	Tianren Ma, Mu Zhang, Yibing Wang, Qixiang Ye Consolidating Reinforcement Learning For Multimodal Discrete Diffusion Models ICLR 2026. [Paper] [Code]
	Tianren Ma, Xiaosong Zhang, Boyu Yang, Junlan Feng, Qixiang Ye ReDDiT: Rehashing Noise for Discrete Visual Generation ICLR 2026. [Paper] [Code]
	Tianren Ma, Lingxi Xie, Yunjie Tian, Boyu Yang, Qixiang Ye ClawMachine: Learning to Fetch Visual Tokens for Referential Comprehension ICLR 2025. [Paper] [Code]
	Tianren Ma, Mingxiang Liao, Xijin Zhang, Qixiang Ye AceTone: Bridging Words and Colors for Conditional Image Grading CVPR 2026. [Paper] [Code]
	Mingxiang Liao, Tianren Ma, Xijin Zhang Open World Image Aesthetic Assessment CVPR 2026 Findings. [Paper] [Code]
	Yunjie Tian, Tianren Ma, Lingxi Xie, Qixiang Ye ChatterBox: Multimodal Referring and Grounding with Chain-of-Questions AAAI 2025. [Paper] [Code]

Co-author Publications

其他论文

	Mu Zhang, Tianren Ma, Yunfan Liu, Kun Hu, Qixiang Ye RebRL: Reinforcing Discrete Visual Diffusion Models with Rebalanced Timestep Credits CVPR 2026. [Paper] [Code]
	Jihao Qiu, Yuan Zhang, Xi Tang, Lingxi Xie, Tianren Ma, Pengyu Yan, David Doermann, Qixiang Ye, Yunjie Tian Artemis: Towards Referential Understanding in Complex Videos* NeurIPS 2024. [Paper] [Code]

* indicates equal contribution.

* 表示作者贡献相同。