Azim Ospanov

prof_pic.jpg

SHB 904

Chung Chi Rd, Ma Liu Shui

Hong Kong, Sha Tin, N.T.

I am a PhD student in Computer Science and Engineering at the Chinese University of Hong Kong, advised by Prof. Farzan Farnia. My research focuses on evaluating and improving generative models, as well as advancing the field of formal mathematics.

I was a member of the first graduating cohort in the “Artificial Intelligence: Systems & Technologies” program at CUHK. Before beginning my PhD, I interned at MetaSystems Asia, working on applied computer vision solutions for healthcare.

My main research interests include:

  • Spectral Theory: Applying spectral methods to generative model evaluation and improving output sampling

  • Formal Mathematics: Using Lean 4 for autoformalization and formal theorem proving

news

Nov 04, 2025 :tada: Our paper on revised version of miniF2F benchmark for Lean4 was accepted at NeurIPS 2025! Link to OpenReview
Nov 04, 2025 :tada: Our paper on repair-driven agent for Lean4 was accepted at NeurIPS 2025! Link to OpenReview
Jun 26, 2025 :tada: Our paper on measuring prompt-aware diversity was accepted at ICCV 2025 as a Spotlight work! Link to the paper
May 07, 2025 :tada: Our paper on statistical convergence of the Vendi scores was accepted at UAI! Link to OpenReview

latest posts

selected publications

  1. APOLLO: Automated LLM and Lean Collaboration for Advanced Formal Reasoning
    Azim Ospanov, Farzan Farnia, and Roozbeh Yousefzadeh
    In The Thirty-ninth Annual Conference on Neural Information Processing Systems, Nov 2025
  2. miniF2F-Lean Revisited: Reviewing Limitations and Charting a Path Forward
    Azim Ospanov, Farzan Farnia, and Roozbeh Yousefzadeh
    In The Thirty-ninth Annual Conference on Neural Information Processing Systems, Nov 2025
  3. Scendi Score: Prompt-Aware Diversity Evaluation via Schur Complement of CLIP Embeddings
    Azim Ospanov, Mohammad Jalali, and Farzan Farnia
    In International Conference on Computer Vision, Oct 2025
  4. Towards a Scalable Reference-Free Evaluation of Generative Models
    Azim Ospanov, Jingwei Zhang, Mohammad Jalali, and 3 more authors
    In The Thirty-eighth Annual Conference on Neural Information Processing Systems, Nov 2024