Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
T5 treats language tasks as text input and text output within one framework.
2020
35,734
Works
Selected research papers and books, linked to their authors.
Training Language Models to Follow Instructions with Human Feedback
InstructGPT uses human feedback to train language models to follow instructions.
2022
33,638
Sequence to Sequence Learning with Neural Networks
Translated sequences with two neural networks, allowing inputs and outputs of different lengths.
2014
33,287
Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
RAG finds relevant documents to help a language model generate better-informed answers.
2020
32,837
GPT-4 Technical Report
Reported GPT-4's capabilities, limitations and safety evaluations as a text-and-image input model.
2023
28,885
Transformers: State-of-the-Art Natural Language Processing
Presented Hugging Face's Transformers library for using, adapting and sharing pretrained language models.
2020
25,911
Language Models are Unsupervised Multitask Learners
GPT-2 showed how learning to predict text could support tasks without separate training.
2019
25,411
Llama 2: Open Foundation and Fine-Tuned Chat Models
Presented Llama 2 and its chat models, including training methods and capability evaluations.
2023
25,241
Qwen3 Technical Report
Presented Qwen3, language models that switch between extended reasoning and quick responses.
2025
21,211
Mistral 7B
Mistral 7B demonstrated strong language-model performance relative to larger models.
2023
15,333