50% on HumanEval with just 1.3B model(twitter.com)

58 points by sytelus 3 years ago | 5 comments

Gzip and KNN Outperforms Transformers on Text Classification(twitter.com)

234 points by rsp1984 3 years ago | 63 comments

XB, or eXtreme Bullshitting: a less misleading name for LLMs(twitter.com)

83 points by gmac 3 years ago | 60 comments

Llama.cpp can do 40 tok/s on M2 Max, 0% CPU usage, using all 38 GPU cores(twitter.com)

40 points by samwillis 3 years ago | 9 comments

Neural networks in the 1990s(twitter.com)

122 points by jrott 3 years ago | 75 comments

Curious about the training data of OpenAI's new GPT-OSS models? I was too(twitter.com)

220 points by flabber 1 year ago | 52 comments

GPT in 500 Lines of SQL(explainextended.com)

1038 points by thunderbong 2 years ago | 70 comments

Blog post from Explain Extended implementing a GPT-style large language model inference in about 500 lines of SQL, as a New Year's post.

Understanding GPT tokenizers(simonwillison.net)

380 points by simonw 3 years ago | 110 comments

Simon Willison blog post explaining how GPT tokenizers convert text to integer tokens, with an Observable notebook tool for encoding, decoding and searching tokens.

GPT-4.1 in the API(openai.com)

659 points by maheshrijal 1 year ago | 484 comments

OpenAI announcement introducing the GPT-4.1 model series in the API, with improvements in coding, instruction following, long context, and a new nano model.

GPT-5 for Developers(openai.com)

445 points by 6thbit 1 year ago | 251 comments

OpenAI's announcement of GPT-5 for developers via its API, pitched as its best model for coding and agentic tasks.

Simply explained: How does GPT work?(confusedbit.dev)

326 points by nitnelave 3 years ago | 279 comments

A blog post from Confused bit giving a simple explanation of how GPT language models like ChatGPT work, starting with their probabilistic approach.

Reproducing GPT-2 in llm.c(github.com)

578 points by tosh 2 years ago | 109 comments

A GitHub discussion in Andrej Karpathy's llm.c repository about reproducing GPT-2 training from scratch in C/CUDA.