rekursiv.ai details how its agent swarm hit SoTA on NanoChat with a 15k-node knowledge graph
hyperparticle · x · 2026-09-15
rekursiv.ai published a full writeup of its auto-autoresearch system: a swarm of AI agents builds on each other's discoveries, preserves failed experiments, and improves how it does research—surpassing the previous NanoChat state of the art in three days.
Setup: improve a language model within a 5-minute training budget on a single B200.
Result: 0.88779 mean bits per byte across 10 seeds, beating the prior best (Recursive 0.91087, Aster 0.90980, Hiloop 0.90160, Hyra 0.90154).
Method:
- Agents treated as "intern savants"—genius knowledge but no sense of what matters—so the swarm figures out the research strategy itself
- A 15,099-node knowledge graph built in 3 days connects issues, artifacts, experiments, beliefs, and code changes
- Built on in-house tools trackinizer, sagent, priml, and configgle
The team previously ran ARC-AGI and Sudoku campaigns and is hiring ([email protected]).
More from coding & agent
- awa-video-reel: free Claude Code skill auto-edits shorts and records website B-roll — TawohAwa · 2026-11-02
- Econ PhD Job Market Database Hits 900+ Candidates, Open to Codex and Claude Code — aniketapanjwani · 2026-11-02
- Cron works by hand, fails on schedule: a 7-point debugging checklist — petrusenko_max · 2026-11-02
- Local quants vs Claude on 9 real private bugs: 3-bit costs thinking time, not score — FeydRowan · 2026-11-02
- Build products for AI, not humans: MCP lets Claude Code run everything — gefei55 · 2026-11-02
- Solo dev launches hosted MCP server with 31 tools scheduling to 20+ social networks — troyizzle · 2026-11-02