Interactive Optimization and LearningAI x Algorithms x Applications

~/pokutta $ whoami

Sebastian Pokutta

Professor @ TU Berlin · Vice President @ Zuse Institute Berlin · Chair @ MATH+ · Chair @ MODAL

# artificial intelligence, optimization, and machine learning

# tl;dr: we use computers to learn from data and make better decisions.

~/pokutta $ cd lab

→ Interactive Optimization and Learning (IOL) iol.zib.de ↗ research ↗ openings ↗

research

My group works on Artificial Intelligence, Optimization, and Machine Learning; from foundational algorithms to real-world systems. We develop new optimization and learning methods with a focus on efficiency and sustainability, combine learning with decision-making, and build agentic AI systems for complex tasks. A particular focus is on AI4Science and AI4Math, where we use AI to drive mathematical discovery and scientific insight. Recent examples include global forest monitoring from satellite data, entanglement certification via optimization, new constructions in extremal combinatorics, AI-assisted mathematical proofs, and AI for creativity and education. group homepage more about research and projects

selected recent work

all publications
  1. 2026
    The KV-Surgeon: Compensation-Aware Offline KV-Cache Compaction Wagner, M., Roux, C., Zimmer, M., and Pokutta, S. preprint arXiv llmmlpruning
  2. 2026
    The exact worst-case asymptotic rate of Nesterov acceleration for strongly convex optimization Pokutta, S., and Stellato, B. preprint arXiv visuals complexitylowerboundsopt
  3. 2026
    CryptHarness: A Harness for Agent-Assisted Cryptanalytic Research Pelleriti, N., Kraus, N., May, A., and Pokutta, S. Proceedings of the NeurIPS 2026 Workshop on Secure and Trustworthy Quantum Machine Learning (SaTQuML) agenticcryptanalysisllmml (Oral Presentation + Workshop)
  4. 2026
    Zero: A Meta-Agent for Mathematical Research Pelleriti, N., Plato, K., Zimmer, M., and Pokutta, S. Proceedings of the Sixth NeurIPS Workshop on Mathematical Reasoning and AI (MATH-AI) agenticai4mathml
  5. 2026
    Curvature-Dependent Lower Bounds for Frank-Wolfe Halbey, J., Roux, C., and Pokutta, S. to appear in Proceedings of NeurIPS arXiv slides complexityfwlowerboundsopt
  6. 2026
    Frank-Wolfe Beyond 1/t Convergence Pokutta, S. to appear in Proceedings of NeurIPS arXiv slides complexityfwopt
  7. 2026
    A Free Lunch in LLM Compression: Revisiting Retraining after Pruning Wagner, M., Roux, C., Zimmer, M., and Pokutta, S. to appear in Proceedings of NeurIPS arXiv llmmlpruningsparsity
  8. 2026
    Within-Model vs Between-Prompt Variability in Large Language Models for Creative Tasks Haase, J., Gonnermann-Müller, J., Hanel, P. H. P., Leins, N., Kosch, T., Mendling, J., and Pokutta, S. to appear in Proceedings of NeurIPS arXiv creativityhaiillmml
  9. 2026
    The Agentic Researcher: A Practical Guide to AI-Assisted Research in Mathematics and Machine Learning Zimmer, M., Pelleriti, N., Roux, C., and Pokutta, S. Proceedings of the ICML 2026 Workshop on AI as a Tool for Mathematics, Computer Science, and Machine Learning (AI4Research) pdf arXiv summary code agenticai4mathml (Oral Presentation + Workshop)
  10. 2026
    Symmetric Extension Complexity of the Spanning Tree Polytope Pokutta, S. preprint arXiv extendedformulationipopt
  11. 2026
    A Counterexample to Ziegler’s Cross-Polytope Conjecture for Simplicial 0/1-Polytopes Kaibel, V., and Pokutta, S. preprint arXiv combinatoricsopt
  12. 2026
    From Associations to Activations: Comparing Behavioral and Hidden-State Semantic Geometry in LLMs Schiekiera, L., Zimmer, M., Roux, C., Pokutta, S., and Günther, F. Proceedings of the 43rd International Conference on Machine Learning (ICML) arXiv summary cognitivellmmlxai
  13. 2026
    Lower Bounds for Frank-Wolfe on Strongly Convex Sets Halbey, J., Deza, D., Zimmer, M., Roux, C., Stellato, B., and Pokutta, S. Proceedings of the 43rd International Conference on Machine Learning (ICML) pdf arXiv summary slides complexityfwlowerboundsopt
  14. 2026
    Neural Concept Verifier: Scaling Prover-Verifier Games via Concept Encodings Turan, B., Asadulla, S., Steinmann, D., Kersting, K., Stammer, W., and Pokutta, S. Proceedings of the 43rd International Conference on Machine Learning (ICML) arXiv summary mlxai (Spotlight + Conference Proceedings)
  15. 2026
    Faster Algorithms for Structured Matrix Multiplication via Flip Graph Search Khoruzhii, K., Gelß, P., and Pokutta, S. Proceedings of the International Symposium on Symbolic and Algebraic Computation (ISSAC) pdf arXiv compalgcomputational (Distinguished Student Author Award at ISSAC 2026)
  16. 2026
    The Hidden Cost of Tokenization: Why (most) Non-English Speakers Pay More for Less Haase, J., and Pokutta, S. preprint arXiv summary fairnessllmmlmultilingual
  17. 2026
    Neural Sum-of-Squares: Certifying the Nonnegativity of Polynomials with Transformers Pelleriti, N., Spiegel, C., Liu, S., Martínez-Rubio, D., Zimmer, M., and Pokutta, S. to appear in Proceedings of the International Conference on Learning Representations (ICLR) arXiv ai4mathcompalgml
  18. 2025
    Conditional Gradient Methods Braun, G., Carderera, A., Combettes, C. W., Hassani, H., Karbasi, A., Mokhtari, A., and Pokutta, S. MOS-SIAM Series on Optimization pdf arXiv mloptsurvey
  19. 2024
    Quantum Optimization: Potential, Challenges, and the Path Forward Abbas, A., Ambainis, A., Augustino, B., Bärtschi, A., Buhrman, H., Coffrin, C., Cortiana, G., Dunjko, V., Egger, D. J., Elmegreen, B. G., Franco, N., Fratini, F., Fuller, B., Gacon, J., Gonciulea, C., Gribling, S., Gupta, S., Hadfield, S., Heese, R., Kircher, G., Kleinert, T., Koch, T., Korpas, G., Lenk, S., Marecek, J., Markov, V., Mazzola, G., Mensa, S., Mohseni, N., Nannicini, G., O’Meara, C., Peña Tapia, E., Pokutta, S., Proissl, M., Rebentrost, P., Sahin, E., Symons, B. C. B., Tornow, S., Valls, V., Woerner, S., Wolf-Bauwens, M. L., Yard, J., Yarkoni, S., Zechiel, D., Zhuk, S., and Zoufal, C. Nature Reviews Physics pdf arXiv optphysicsquantumsurvey
  20. 2024
    The Frank-Wolfe algorithm: a short introduction Pokutta, S. Jahresbericht der Deutschen Mathematiker-Vereinigung, 126, 3–35 pdf arXiv mlopt

talks & teaching

archive
help.txt · keyboard shortcuts

navigate

1 … 6go to tab (home, bio, publications, blog, archive, iol lab)
gt / gTnext / previous tab
j / kscroll down / up
d / uhalf a page down / up
gg / Gtop / bottom of the page
]] / [[next / previous section

search and commands

/textsearch this page, n / N for next / previous match; on publications it filters the list (:noh shows all again)
:command line, e.g. :bio :pubs :blog :lab :set bg=dark :noh :help (Tab completes)
:eopen this page in vim (pretend editing; :q closes it)
:newnew note in a full-window editor (stays in this browser; :q returns here, :notes lists them)
Escclose, clear search

other

ttoggle dark mode
itype in the prompt (home; on desktop it starts active), e.g. help ls papers grep wolfe · Esc switches to the keys above
pflip the avatar (home)
?this help
shortcuts are inactive while typing in a field