Large Language Models as Markov Chains(arxiv.org)

63 points by Anon84 1 year ago | 33 comments

Are large language models a threat to digital public goods?(arxiv.org)

118 points by oss_fan 3 years ago | 117 comments

Large language models in national security applications(arxiv.org)

74 points by bindidwodtj 1 year ago | 38 comments

Large Language Models Are Neurosymbolic Reasoners(arxiv.org)

94 points by optimalsolver 2 years ago | 86 comments

A PhD student's perspective on research in NLP in the era of LLMs(arxiv.org)

96 points by morgangiraud 3 years ago | 35 comments

PhD Knowledge Not Required: A Reasoning Challenge for Large Language Models(arxiv.org)

166 points by enum 1 year ago | 78 comments

Language models can explain neurons in language models(openai.com)

659 points by mfiguiere 3 years ago | 416 comments

OpenAI research post describing using GPT-4 to automatically generate and score explanations of individual neuron behavior in GPT-2.

AI tools are spotting errors in research papers(nature.com)

555 points by kgwgk 1 year ago | 200 comments

Nature news article on a growing movement using AI and large language models to detect errors in research papers, sparked by a flawed black plastic utensils study.

Updated practice for review articles and position papers in ArXiv CS category(blog.arxiv.org)

483 points by dw64 11 months ago | 228 comments

arXiv blog post announcing that its computer science category has updated moderation practice for review/survey articles and position papers.

Training LLMs from ground zero as a startup(yitay.net)

360 points by swyx 2 years ago | 92 comments

Yi Tay's blog post on the experience of training large language models from scratch at a startup (Reka), covering compute, infrastructure, and lessons learned.

Deep Learning Is Not So Mysterious or Different(arxiv.org)

445 points by wuubuu 1 year ago | 112 comments

arXiv paper by Andrew Gordon Wilson arguing deep learning generalization phenomena like benign overfitting and double descent are explainable by soft inductive biases.

Fine tune a 70B language model at home(answer.ai)

821 points by jph00 2 years ago | 197 comments

Answer.AI blog post announcing an open source system combining FSDP and QLoRA to train a 70B language model on two 24GB consumer GPUs.