Home›Papers
Papers
Papers
Papers that changed AI, with a plain-language explainer for each.
Total 100 · Clear filters
- 2019
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer (T5)
JMLR
A 2019 Google paper that frames every language task as text in, text out, using one encoder-decoder Transformer and one training setup. It also introduced the C4 dataset, and the released T5 models became common research starting points.
- 2019
RoBERTa: A Robustly Optimized BERT Pretraining Approach
arXiv
A replication study of BERT pretraining that measures how much key hyperparameters and training data size matter, finds BERT was significantly undertrained, and shows a properly trained BERT matches or beats later models, reaching state of the art on GLUE, RACE and SQuAD.
- 2019
XLNet: Generalized Autoregressive Pretraining for Language Understanding
arXiv
XLNet is a generalized autoregressive pretraining method from a 2019 paper. It learns bidirectional contexts by maximizing expected likelihood over all factorization-order permutations and outperforms BERT on 20 tasks under comparable settings.
- 2018
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
NAACL
Google's 2018 BERT paper introduced an encoder-only Transformer that reads text in both directions and is pre-trained by hiding words. It set new results on eleven language tasks and made download-and-fine-tune the default workflow.
- 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
Improving Language Understanding by Generative Pre-Training (GPT)
OpenAI
OpenAI's 2018 paper introducing GPT, a Transformer decoder pre-trained to predict the next token and then fine-tuned for specific tasks. It improved the state of the art on 9 of 12 tasks studied and set the recipe the GPT series followed.
- 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.
- 2018
Deep contextualized word representations
arXiv
A 2018 paper by Matthew E. Peters and five co-authors that represents each word with the internal states of a pre-trained deep bidirectional language model, so vectors change with context; adding them to existing models improved results on six NLP tasks.
- 2018
Universal Language Model Fine-tuning for Text Classification
arXiv
ULMFiT is a 2018 transfer-learning method by Jeremy Howard and Sebastian Ruder for fine-tuning language models across NLP tasks; it beat the state of the art on six text-classification tasks and performed well with few labels.
- 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.
- 2017
Focal Loss for Dense Object Detection
arXiv
A 2017 paper by Lin, Goyal, Girshick, He and Dollar that argues dense one-stage detectors are held back by extreme foreground-background imbalance and proposes focal loss, a reshaped cross-entropy that down-weights well-classified examples.
- 2017
Proximal Policy Optimization Algorithms
arXiv
A 2017 paper by Schulman, Wolski, Dhariwal, Radford and Klimov that introduced proximal policy optimization (PPO), a reinforcement learning method that is much simpler to implement and tune and became OpenAI's default RL algorithm.
- 2017
Attention Is All You Need
NeurIPS
A 2017 paper that introduced the Transformer, a neural network built entirely on attention with no recurrence. It set new machine translation results and became the architecture behind almost every large language model since.
- 2017
Mask R-CNN
arXiv
Mask R-CNN is a 2017 framework for object instance segmentation. It detects objects and produces a high-quality mask for each instance by adding a parallel mask branch to Faster R-CNN.
- 2016
Neural Architecture Search with Reinforcement Learning
arXiv
A 2016 paper by Barret Zoph and Quoc V. Le that uses a recurrent network trained with reinforcement learning to design neural network architectures automatically, matching human-designed models on CIFAR-10 and beating LSTM on Penn Treebank.
- 2016
Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
arXiv
A 2016 paper describing Google's neural machine translation system, GNMT: a deep LSTM encoder-decoder with attention, designed to make end-to-end translation practical by addressing rare words, parallelism and inference cost.
- 2016
WaveNet: A Generative Model for Raw Audio
arXiv
WaveNet is a deep neural network that generates raw audio waveforms one sample at a time, predicting each sample from all the ones before it. Used for text-to-speech, it was rated by listeners as more natural than the systems it was compared with.
- 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.