Formalizations of theorems from papers submitted to ACM STOC 2027 (the 59th Annual ACM Symposium on Theory of Computing). Paper list: https://sigact-stoc.github.io/stoc2027/ . Missions here state a main result of a STOC 2027 paper in Lean 4 with Mathlib; each mission's title begins with the paper title.
Collections
Curated groups of missions around a topic, textbook, or research area.
- 11%
A coordinated effort to formalize the mathematics of operations research in Lean 4.
Scope: missions in linear and integer programming, convex and nonlinear optimization, combinatorial optimization, dynamic programming, scheduling, stochastic models such as queueing and inventory, simulation, revenue management, and game theory and mechanism design.
73%A coordinated effort to formalize the mathematics of quantum information and quantum computation in Lean 4.
Scope: missions in quantum information theory, quantum computation and algorithms, quantum complexity theory, quantum error correction, entanglement theory, and quantum learning.
33%The TCS Formalization Drive
Almost thereA coordinated effort to formalize the mathematics of theoretical computer science in Lean 4.
Scope: missions in computational complexity, algorithms and optimization, data structures, automata and computability, randomized and approximation algorithms, cryptography, learning theory, information theory, algorithmic game theory, combinatorics, computational geometry, and distributed computing.
85%Named conjectures and open problems with a precise Lean statement, from Riemann and Goldbach to Collatz and the Jacobian conjecture.
29%A collective effort to advance our understanding of the generalized Hodge conjecture. Let's work on explicit constructions for an important subcase: sextic fourfolds with an involution. This case was isolated by Voisin; a reduction carried out by Diamond makes its prediction computationally effective. See more at hodge.computer.
If you manage to complete a mission, please mirror your work back to hodge.computer. Let's get to work!
0%Problems from the Erdős problem list, each formalized as its own mission. Settle one, or decompose it into lemmas.
45%A collective effort to tackle ten open problems in combinatorics, formalized in Lean 4. Each challenge is scaled by a parameter : small values are known results, larger ones are open, and every mission here lists the known rungs and the first open one as milestones. Let's prove new cases and push the known bounds forward. The Lean definitions are those of the public Ten Challenges repository, so a proof of a goal or of any rung produced here is also welcome on:
where accepted submission appear in the leaderboard (with public and private tracks).
0%Queueing and Stochastic Networks
Almost thereSingle-server queues, Jackson and loss networks, heavy-traffic limits, fluid stability, and the control of queueing systems.
88%Scheduling Theory
Almost thereMachine, flowshop, jobshop and project scheduling: optimality of classic rules, complexity reductions, and approximation guarantees.
94%OpenAI Ten Proofs
CompleteTen missions drawn from the OpenAI Ten Proofs formalizations, one per proof. Each mission uses the original Comparator targets: one main goal, with the remaining targets as milestones. The Lean statements and proofs use Lean 4.33.1.
100%- 100%
OAI Math
CompleteAll of openAI's results.
100%Bertsimas and Tsitsiklis's Introduction to Linear Optimization: polyhedra, the simplex method, duality, and the ellipsoid method.
100%Bandit Algorithms
CompleteLattimore and Szepesvári's Bandit Algorithms: regret bounds for explore-then-commit, UCB, Thompson sampling, and adversarial bandits.
100%Convex Polytopes
CompleteGrünbaum's Convex Polytopes: convex representations, face lattices, facet growth, and reconstruction from partial data.
100%Markov Decision Processes
CompleteBäuerle and Rieder's Markov Decision Processes with Applications to Finance: Bellman equations, optimal policies, partial observation, and optimal stopping.
100%Understanding Machine Learning
CompleteShalev-Shwartz and Ben-David's Understanding Machine Learning: PAC learning, VC dimension, and the fundamental theorem of statistical learning.
100%Convex Optimization
CompleteConvex optimization textbooks, chapter by chapter: KKT conditions, conic duality, barrier methods, and the complexity of first-order methods from center of gravity to mirror descent.
100%Vershynin's High-Dimensional Probability and Wainwright's High-Dimensional Statistics: concentration, random matrices, and sparse recovery.
100%