Mathematics
7 verified cases, highest-rated first.
Industries: scientific computing 2, automated mathematics 1, mathematics 1, AI research 1, machine learning methods 1, operations research 1
-
Solving olympiad geometry without human demonstrations
Automated proving of olympiad-level geometry theorems was blocked by the scarcity of human demonstrations in machine-verifiable format.
Mathematics · 2024 · solved · 7/10 importance
-
Discovering faster matrix multiplication algorithms with reinforcement learning
Discovering more efficient matrix multiplication algorithms is hard because the space of possible algorithms is enormous and human intuition is limited.
Mathematics · 2022 · partially solved · 7/10 importance
-
Mathematical discoveries from program search with large language models
LLM confabulations hinder scientific discovery, and the cap set problem in extremal combinatorics lacked constructions beyond the best-known ones.
Mathematics · 2023 · solved · 6/10 importance
-
Superhuman AI for multiplayer poker
AI had surpassed humans in two-player games, but six-player no-limit Texas hold'em poker remained a much harder, unsolved challenge.
Mathematics · 2019 · solved · 6/10 importance
-
Symbolic regression as a feature engineering method for machine and deep learning regression tasks
Manually designing input features for machine and deep learning regression models relies on domain expertise and limits predictive performance.
Mathematics · 2024 · solved · 4/10 importance
-
Efficient generator of mathematical expressions for symbolic regression
Symbolic regression — finding mathematical equations that fit data — is computationally hard.
Mathematics · 2023 · partially solved · 4/10 importance
-
Supplementing recurrent neural networks with annealing to solve combinatorial optimization problems
Simulated annealing suffers slow convergence and gets stuck in local search spaces when solving combinatorial optimization problems like Max-Cut, nurse scheduling, and TSP.
Mathematics · 2023 · ai assisted · 4/10 importance