Self-Rewarding LLMs isn't news: RLAIF has long been standard practice
burny_tech · x · 2026-09-20
A viral post framed Meta's 2024 ICML paper "Self-Rewarding Language Models" as a breakthrough toward recursive self-improvement; burnytech points out RL from AI feedback has been standard practice for a while. The paper uses LLM-as-a-Judge to provide its own rewards during iterative DPO: three iterations on Llama 2 70B beat Claude 2, Gemini Pro, and GPT-4 0613 on AlpacaEval 2.0.
More from Fun
- DeepMind researcher quips we're solving intelligence with the dumbest method imaginable — wgrathwohl · 2026-11-12
- "Some Academics Are Even Proud to Have Never Tried AI" — LLM Debate Flares Again — FlorianGallwitz · 2026-11-12
- A Swarm of Claudes Built Its Own Video Editor and Cut NASA Footage Overnight — zealcaiden · 2026-11-12
- Worst AI use: signaling you don't care about things you should — Afinetheorem · 2026-11-12
- LLM controls vocal synth to perform Ave Maria in 4-part harmony — AAAzzam · 2026-11-12
- AI naming meme: LateOn and DenseOn hailed as embedding model name masterpieces — antoine_chaffin · 2026-11-12