AI agent swarm beats NanoChat benchmark SoTA in 3 days with 15k-node knowledge graph
hyperparticle · x · 2026-09-15
rekursiv.ai sent a swarm of AI agents at Karpathy's NanoChat benchmark and surpassed the previous state of the art in three days.
The team built a custom harness on top of a graph database for "auto-autoresearch," letting each iteration learn from failed experiments, and logged over 15,000 knowledge-graph entries in the process.
More from coding & agent
- Tencent unveils Hyra 2.0, a research agent that builds reusable scientific tools — TencentHunyuan · 2026-11-02
- General Reasoning explains how to train agent swarms with RL and latency pressure — rosstaylor90 · 2026-11-02
- "Coding isn't the bottleneck anymore — usage windows are," says founder Kyle Gawley — kylegawley · 2026-11-02
- Demo passed ≠ production-ready: 5 questions to answer before letting agents run your business — _kellhus · 2026-11-02
- An AI agent passed 100 test chats—still not ready to run the business — djgoel123 · 2026-11-02
- Agent given total creative freedom produces a 1:44 code-drawn animated short film — ivegotnoidea1 · 2026-11-02