About

Coding, Life, and AGI

Interests

Teaching AI how to reason about the world

Recent Highlight: SambaNova's Llama-3.1-Instruct-O1

I built SambaNova's recent work challenging OpenAI's O1 model with a Llama 3.1-powered demo. Check out our implementation on HuggingFace:

Llama3.1-Instruct-O1 Demo

Read more about this development in the VentureBeat article.

Selected Publications

Memory-Efficient LLM Training with Online Subspace Descent [NeurIPS 2024]

Kaizhao Liang, Bo Liu, Lizhang Chen, Qiang Liu

Pixelated Butterfly: Simple and Efficient Sparse training for Neural Network Models [ICLR 2022]

Beidi Chen, Tri Dao, Kaizhao Liang, Jiaming Yang, Zhao Song, Atri Rudra, Christopher Re

Uncovering the Connections Between Adversarial Transferability and Knowledge Transferability [ICML 2020]

Kaizhao Liang*, Jacky. Y. Zhang*, Boxin Wang, Zhuolin Yang, Sanmi Koyejo, Bo Li

Relevant Projects

Gesture Control of Autonomous Vehicle

A new interface for human interaction with Autonomous Vehicles.

Interval Bound Propagation PyTorch

Exploring certified defense against adversarial attacks with limited resources by using a "tighter" objective.

Review Service

CVPR2021, ICCV2021, ICLR2022, CVPR2022, AISTATS2022, NEURIPS2023, ICLR2023, CVPR2023

Contact

kaizhao.liang@sambanovasystems.ai

kaizhaol@utexas.edu