Welcome, it’s a pleasure to connect with you! I’m Bingcong Li, a postdoctoral researcher at ETH Zurich collaborating with Prof. Niao He and the ODI group. Prior to this, I received doctoral degree from the University of Minnesota under the supervision of Prof. Georgios B. Giannakis, and then I gained industry experience, dedicating a year to LLMs.

Find me via bingtsongli[@]gmail.com or bingcong.li[@]inf.ethz.ch.

General interests

Computation is at the heart of modern AI and the invisible engine behind the success of LLMs. The overarching goal of my research is to make computation in the era of LLMs efficient, accessible, and affordable across heterogeneous resources, from clusters with thousands of GPUs to users with consumer-grade hardware. My research draws on interdisciplinary tools from deep learning, optimization, and signal processing to tackle challenges across multiple scales of computation. At every stage, I develop theoretically grounded methods so that the resulting systems are more explainable. I focus on three questions:

• Foundations of computing: Which architectures and training strategies deliver the best accuracy and efficiency?
• Scaling computation up: How do we sustain throughput, reliability, and cost efficiency at the scale of thousands of GPUs?
• Personalization: How can we empower individuals and organizations with limited resources to access and benefit from modern AI?

I enjoy cycling 🚴🏻 outside offices. I also do a bit gym training, but ocationally people tell me my triceps pushdown techniques could use some work.

Recent updates

• 10/2025. We are hosting the Efficient LLMs Fine-tuning (ELF) Track in AI+X summit. See you in October, Zurich!
• 07/2025. I will talk about Riemannian optimization and its provable merits for fine-tuning LLMs in EUROPT 2025.
• 06/2025. I will talk about “LoRA sugery” at Efficient Machine Learning Reading Group.
• 06/2025. [New paper] LoRA does not use allocated rank effectively. This can be addressed with PoLAR, a co-design of architecture and optimizer. Check out our paper.
• 06/2025. [New paper] RefLoRA optimally rescales/refactorizes LoRA per training step to make fine-tuning LLMs faster. Check out our paper.
• 06/2025. [New paper] Zeroth-order methods provably find flat minima. Check it out here.
• 05/2025. [ICML 2025] Transfer learning provably benefits RLHF. Check out our paper.
• 04/2025. Talked about “Fine-tuning LLMs cost-efficiently” at Peking University.
• 01/2025. [ICLR 2025] We prove that initialization exponentially impacts the convergence behavior of ScaledGD on LoRA type problems (i.e., linear –> quadratic rates).
• 12/2024. Talked about “Architecture-Aware Optimization” at ELLIS UnConference.
• 12/2024. [ICASSP 2025] A new variant of SAM is released.
• 09/2024. [NeurIPS 2024] We study the implicit regularization of sharpness-aware minimization (SAM) and explicify it to alleviate computational burdern of SAM. The resultant approach is useful for finetuning LLMs with LoRA.
• 05/2024. [ICML 2024] Memory-efficient private finetuning for LLMs. We also have a paper at Theoretical Foundations of Foundation Models (TF2M) workshop.
• 01/2024. Start as a postdoc in ETH Zurich, working with Prof. Niao He.
• 12/2023. [ICASSP 2024] Universal ‘preconditioner’ for meta learning.
• 09/2023. [NeurIPS 2023] Improving generalization by refining optimization of sharpness-aware minimization; see here.