Back to timeline

Research · March 13, 2023

Alpaca

Stanford fine-tuned LLaMA on 52 thousand examples generated by GPT-3.5 itself and got InstructGPT-like behaviour for 600 dollars.

Why it matters

Alignment that cost OpenAI hundreds of raters was reproduced for the price of a laptop, using the larger model's own output.

The technique is called distillation: a larger model generates training examples for a smaller one. Legally this conflicted with OpenAI's terms of use, and the question stayed open. Practically it meant that an advantage bought with expensive alignment does not hold: it can be siphoned. Stanford later took the demonstration down over safety and cost.

Event record

Event date
March 13, 2023
Timeline date
Event date
Verification
Sources gathered automatically · September 17, 2026
Lines
ID
evt-0272

Announcement of 13 March 2023.

Sources

Related events

Records that link to this one