Tencent's Hy3 Model Solves 50-Year-Old Combinatorics Problem
Tim_Dettmers · x · 2026-07-30
Tencent's Hunyuan team utilized its research agent Hyra and the Hy3 model to solve a combinatorics problem that had been open for over half a century.
The question explored how much faster the sumset of a finite set of integers can grow compared to its difference set. A 1969 theorem established an upper bound of 2 for the exponent, but for over 50 years, the best constructions (including recent AI-assisted attempts by Terence Tao and others) barely exceeded 1.1.
Assisted by the Hy3 model, researchers found an elegant, explicit construction achieving a lower bound of 2, perfectly matching the upper bound and completely resolving the problem.
Related event: Tencent Hunyuan AI Solves 50-Year-Old Combinatorics Problem(3 posts)→
More from Models
- Reflection's Beam, a 501B-parameter open MoE model with 1M context, lands on Hugging Face under Apache 2.0 — QGallouedec · 2026-10-16
- Why are sub-15B open models so rare now? Labs seem to have settled on 24-27B minimum — Hot_Example_4456 · 2026-10-16
- User spots model openly generating "ideal responses" to train other models — liminal_bardo · 2026-10-16
- Reflection's 501B MoE Beam hits 463 tok/s on 8x GB200 with NVFP4 at day 0 — zhyncs42 · 2026-10-16
- Gemma 4 31B QAT on one 24GB card: 46 tok/s plain, ~80 with draft model on RTX 4090 — Subject_Poet_3861 · 2026-10-16
- JevBench v1.7.46 leaderboard: new d1-ultra tops 32 API systems at 76.1 composite — airesearch12 · 2026-10-16