Researchers at Stanford took down their Alpaca AI, a short-lived chatbot that harnessed Meta’s LLaMA AI. Alpaca went public last week, and came down shortly as costs went up and safety risks became more apparent.
So far, researchers can't get AI to behave. The Stanford model, based on Meta's LLaMA AI, went down as fast as it went up
October 1, 2022: Volume XC, No. 19 by Kirkus Reviews - Issuu
Summary of the State of AI Report 2023 - by Michael Spencer
Stanford Axes Alpaca AI, Facebook LLama Copy, Over Cost, Errors
RLHF 201 - with Nathan Lambert of AI2 and Interconnects
anton on X: Anyone try reproducing stanford's LLaMA Alpaca? I haven't seen loss curves like this. I reproduced it against another decoder model My attempt is on the left (it doesn't use
Game Changing AI from Stanford - Alpaca
Run Stanford Alpaca 7B LLAMA Instruction Finetuned Model on Local System, Dalai
Multimodal LM roundup: Unified IO 2, inputs and outputs, Gemini, LLaVA-RLHF, and RLHF questions
Do By Friday
How does Alpaca follow your instructions? Stanford Researchers Discover How the Alpaca AI Model Uses Causal Models and Interpretable Variables for Numerical Reasoning - MarkTechPost
Information, Free Full-Text
Product & Engineering Archives - Pear VC
Stanford Researchers Take Down Alpaca AI Due to 'Hallucinations' and Rising Costs
Stanford CRFM
New LLM Foundation Models - by Sebastian Raschka, PhD