Schedule
All discussion classes center around a required reading. The schedule also includes relevant papers not cited by those papers (since the Archeologist’s job is to uncover relevant papers cited by the required reading, and we don’t want to make their job too easy!) as well as some relevant papers by myself that I’m always happy to discuss during office hours or after class.
Foundations lectures
Algorithmic reasoning with transformers and LLMs
- 10/7
- In-Context Learning of Regression Algorithms
- Discussion
- Required reading: Akyürek, Ekin, et al. “What learning algorithm is in-context learning? investigations with linear models.” ICLR’23. [link]
- Additional resources:
- 10/9
- How Transformers Select Algorithms In-Context
- Discussion
- Required reading: Bai, Yu, et al. “Transformers as statisticians: Provable in-context learning with in-context algorithm selection.” NeurIPS’23. [link]
- Additional resources:
- Extended version of the paper
- Talk by Song Mei
- Xu, Lin, et al. “SATzilla: portfolio-based algorithm selection for SAT.” Journal of Artificial Intelligence Research 32 (2008): 565-606. [link]
- 10/14
- Algorithmic Understanding of GenAI
- Discussion
- Required reading: Eberle, Oliver, et al. “Position: We Need An Algorithmic Understanding of Generative AI.” ICML’25. [link]
- 10/16
- Auto-Designing Heuristics with LLMs
- Discussion
- Required reading: Liu, Fei, et al. “Evolution of Heuristics: Towards Efficient Automatic Algorithm Design Using Large Language Model.” ICML’24. [link]
- Additional resources:
- Streeter, Matthew, et al. “Combining multiple heuristics online.” AAAI’07. [link]
Algorithmic reasoning with graph neural networks
- 10/21
- GNNs as CSP Heuristics
- Discussion
- Required reading: Tönshoff, Jan, et al. “One model, any CSP: graph neural networks as fast global search heuristics for constraint satisfaction.” IJCAI’23. [link]
- Additional resources:
- Horvitz, Eric, et al. “A Bayesian approach to tackling hard computational problems.” AAAI’01. [link]
- 10/23
- Neural Algorithmic Reasoning
- Discussion
- Required reading: Numeroso, Danilo, Davide Bacciu, and Petar Veličković. “Dual algorithmic reasoning.” ICLR’23. [link]
- 10/28
- Introduction to diffusion models
- Lecture
- 10/30
- Graph Diffusion for Combinatorial Optimization
- Discussion
- Required reading: Sun, Zhiqing, and Yiming Yang. “Difusco: Graph-based diffusion solvers for combinatorial optimization.” NeurIPS’23. [link]
- 11/4
- No class
Mathematical optimization
- 11/6
- LLMs for Optimization Modeling
- Discussion
- Required reading: AhmadiTeshnizi, Ali, et al. “OptiMUS-0.3: Using large language models to model and solve optimization problems at scale.” arXiv’25. [link]
- Additional resources:
- Lawless, Connor, et al. “LLMs for cold-start cutting plane separator configuration.” CPAIOR’25. [link]
- 11/11
- Predict-and-Search for Integer Programming
- Discussion
- Required reading: Huang, Taoan, et al. “Contrastive predict-and-search for mixed integer linear programs.” ICML’24. [link]
- Additional resources:
- Balcan, Maria-Florina, et al. “Learning to branch: Generalization guarantees and limits of data-independent discretization.” Journal of the ACM 71.2 (2024): 1-73. [link]
- 11/13
- Differentiable Integer Programming
- Discussion
- Required reading: Geng, Zijie, et al. “Differentiable integer linear programming.” ICLR’25. [link]
- Additional resources:
- Wang, Po-Wei, et al. “SATnet: Bridging deep learning and logical reasoning using a differentiable satisfiability solver.” ICML’19. [link]
Theoretical guarantees
- 11/18
- Policy Gradients for Solution Samplers
- Discussion
- Required reading: Caramanis, Constantine, et al. “Optimizing solution-samplers for combinatorial problems: The landscape of policy-gradient method.” NeurIPS’23. [link]
- 11/20
- Approximation Algorithms with Predictions
- Discussion
- Required reading: Antoniadis, Antonios, et al. “Approximation algorithms for combinatorial optimization with predictions.” ICLR’25. [link]
- Aditional resources:
- Shen, Judy Hanwen, Ellen Vitercik, and Anders Wikum. “Algorithms with calibrated machine learning predictions.” ICML’25. [link]
- 11/25
- No class
- 11/27
- No class
Final project
- 12/2
- Guest lecture
- 12/4
- Final project presentations