Bonsai 2 quant quietly mauled performance, independent tests confirm
julianharris · x · 2026-09-20
A viral claim that Bonsai 2 retains 98%+ of its intelligence at only 5.9 GB doesn't hold up: julianharris notes the quantization badly hurt performance, a detail buried deep in the team's own paper and now independently validated as "pretty hopeless." Bonsai 1 had similar issues. The "frontier intelligence at home" hype collides with real-world results.
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