Home›Papers
Papers
Papers
Papers that changed AI, with a plain-language explainer for each.
Total 49 · Clear filters
- 2026
GLM-5: from Vibe Coding to Agentic Engineering
arXiv
Zhipu's GLM-5 report describes a 744-billion-parameter mixture-of-experts model built on sparse attention and asynchronous reinforcement learning, aimed at moving AI-assisted programming from vibe coding toward agentic engineering.
- 2026
Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces
arXiv
Terminal-Bench 2.0 is a curated benchmark of 89 hard tasks carried out in computer terminal environments, built to measure how well AI agents handle realistic command-line work; frontier models and agents score below 65% on it.
- 2024
FrontierMath: A Benchmark for Evaluating Advanced Mathematical Reasoning in AI
arXiv
FrontierMath is a benchmark of hundreds of original, unpublished mathematics problems written and vetted by expert mathematicians, introduced by Epoch AI to test advanced mathematical reasoning in AI. At release, state-of-the-art models solved under 2% of the problems.[^1][^2]
- 2024
Accurate structure prediction of biomolecular interactions with AlphaFold 3
Nature
AlphaFold 3, published in Nature in May 2024, extends structure prediction to biomolecular interactions, reporting greater accuracy than docking tools for protein–ligand binding and than AlphaFold-Multimer v.2.3 for antibody–antigen complexes.
- 2024
LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code
arXiv
A benchmark for code LLMs that keeps collecting fresh contest problems from LeetCode, AtCoder and CodeForces, so that scores reflect ability rather than memorised test data, and that also tests code repair and execution reasoning.
- 2023
Scaling deep learning for materials discovery
Nature
A 2023 Nature paper from researchers at Google DeepMind presenting a scaled-up deep-learning approach to predicting stable crystal structures, screening 2.2 million candidates and reporting a large jump in stability-prediction accuracy.
- 2023
GPQA: A Graduate-Level Google-Proof Q&A Benchmark
arXiv
GPQA is a benchmark of 448 expert-written multiple-choice questions in biology, physics and chemistry, built so that skilled non-experts cannot look the answers up, and hard enough that the strongest GPT-4 baseline scored 39% while domain experts reached 65%.
- 2023
SWE-bench: Can Language Models Resolve Real-World GitHub Issues?
arXiv
A benchmark of 2,294 real software engineering problems taken from GitHub issues and the pull requests that resolved them, in which a model must edit a codebase to fix an issue; on release the best model solved under 2% of them.
- 2022
Learning skillful medium-range global weather forecasting
Science
GraphCast is a graph neural network trained on four decades of reanalysis data that produces 10-day global weather forecasts at 0.25 degree resolution in under a minute, outperforming the leading operational deterministic systems on most verification targets.
- 2022
Magnetic control of tokamak plasmas through deep reinforcement learning
Nature
A 2022 Nature paper by Degrave, Felici and colleagues showing that a deep reinforcement learning controller can learn on its own to command all of a tokamak's magnetic control coils, producing and holding a range of plasma shapes.
- 2021
Program Synthesis with Large Language Models (MBPP)
arXiv
A 2021 arXiv paper by Austin, Odena, Nye and eight other authors that introduced MBPP, a benchmark of 974 short Python programming tasks, and measured how language models from 244M to 137B parameters turn natural-language descriptions into working code.
- 2021
Highly accurate protein structure prediction with AlphaFold
Nature
The Nature paper described AlphaFold's methods, system design, and evaluation for protein-structure prediction.
- 2021
Evaluating Large Language Models Trained on Code (Codex / HumanEval)
arXiv
A 2021 paper from OpenAI introduces Codex, a GPT model fine-tuned on public GitHub code, and HumanEval, an evaluation set that judges generated Python programs by whether they work rather than by how closely they match a reference.
- 2019
Mastering Atari, Go, chess and shogi by planning with a learned model
Nature
MuZero is a DeepMind reinforcement-learning algorithm that learns a model of its environment and plans inside that model, mastering Atari, Go, chess and shogi without being given the rules of those games.
- 2018
Neural Ordinary Differential Equations
arXiv
A 2018 paper by Chen, Rubanova, Bettencourt and Duvenaud that replaces a fixed sequence of hidden layers with a neural network describing the derivative of the hidden state, so outputs are computed by an ODE solver — yielding continuous-depth models with constant memory cost.[^1]
- 2018
Group Normalization
arXiv
Group Normalization is a 2018 paper by Yuxin Wu and Kaiming He that proposes normalizing channels in groups, independent of batch size, to address batch normalization's weakness when batches are small.
- 2017
A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play
Science
A 2017 arXiv paper, later published in Science, reporting that a single self-play reinforcement learning algorithm, AlphaZero, mastered chess, shogi and Go and beat strong specialised programs in each.
- 2017
mixup: Beyond Empirical Risk Minimization
arXiv
mixup is a training principle from a 2017 paper by Zhang, Cisse, Dauphin and Lopez-Paz that trains networks on convex combinations of pairs of examples and their labels, improving generalisation on image, speech and tabular benchmarks.
- 2017
Mastering the game of Go without human knowledge
Nature
AlphaGo Zero learned Go only by playing against itself, starting from random play, and beat the earlier champion-defeating AlphaGo 100 games to 0 after three days of training. The Nature paper appeared in October 2017.
- 2016
Identity Mappings in Deep Residual Networks
arXiv
A 2016 analysis of residual networks arguing that with identity skip connections and the activation placed after addition, signals travel directly between blocks; a new residual unit built on this eases training, and a 1001-layer network reaches 4.62% error on CIFAR-10.