GPT-5.5 Solves Open Math Problem
khademinori · x · 2026-07-12
A repost highlights that, with prompt guidance, GPT-5.5 quickly solved a well-known open problem in discrete mathematics: the optimal bound for oblivious online vector balancing.
The original post emphasized a key insight: the author often found that their long-held "perspective" on such problems was entirely wrong. The solution provided by the model approached the issue through the lens of "the sum of 3 Gaussian distributions," making it much more tractable. The poster, who spent significant time on this problem during their postdoc, found this paradigm shift highly inspiring.
More from Models
- Rumor claims Kimi faked performance by serving Claude; DeepSeek new model surprises in evals — realsohamparekh · 2026-09-11
- GPT-5.6 writes well but is instantly forgettable, user complains — BasedRaddka · 2026-09-11
- Opus Refuses Protein Research Codebase Over 'Safety' Concerns, Dev Considers Rolling His Own — josephdviviano · 2026-09-11
- User Hails Unconfirmed 'DeepSeek 4.1 Flash' as an Inflection Point in LLMs — himanshustwts · 2026-09-11
- Terminal Bench v4: GLM-5.3 Leads at 41.9%, Kimi-K3 Underwhelms at 12.6% — Ok_Warning2146 · 2026-09-11
- GPT-6 Astra beats Factorio with enemies in 44 in-game hours at ~$4,500 API cost — liminal_bardo · 2026-09-11