Stories by swyx

FireFunction V1 – GPT-4-level function calling model – 4x faster, open weights

Unexpected responses from ChatGPT: Incident Report

Groq's low pricing is only possible with high batch size

Tasting Notes from Evil Rabbit

Representation Engineering Mistral-7B an Acid Trip

Worthwhile Research for Building SOTA LLMs

Suppressing Pink Elephants with Direct Principle Feedback

All Neural Networks, All Autonomous, All 1X Speed

Cloud Intelligence at the speed of 5000 tok/s – with Together AI

Starter Guide to SF for Founders

Scale is all you need

Why StackOverflow usage is down 50%

Self-rewarding-lm-PyTorch: Self-Rewarding Language Model from MetaAI

The Sliced-Crosswise Only-On-Tuesday World

Add Whisper to Your iOS Action Button

I Reverse Engineered Vercel's v0.dev Prompt and Code Optimization Logic

Preference Tuning LLMs with Direct Preference Optimization Methods

Sam Altman at Davos [video]

How to train your own Large Multimodal Model

Four Magic Words

7
8
9
10
11
12
13
14
15
16