Meituan LongCat team's Data World synthesizes data environments to train data agents
meituan-longcat · hf · 2026-10-16
Meituan's LongCat Interaction Team, with Nanjing University and other institutions, released Data World: a multi-agent engine that synthesizes data environments and tasks from scratch to train models for analytical insight discovery.
- Motivation: curating real environments for data agents is costly and constrained by privacy and IP restrictions
- The approach scales environments via exploration-based synthesis of seeds, tasks, and rubrics, without trajectory distillation
- The paper reports improved performance across benchmarks
More from Research
- A beginner-friendly introduction to causal inference kicks off, Part 1 — Nasereliver · 2026-10-16
- FBI method trains for Pass@K with one sample per update at 12.3% of the cost — EmtiyazKhan · 2026-10-16
- Berkeley researchers sample humanoid robot policies directly from weight space with diffusion — berkeley_ai · 2026-10-16
- OneWM: unified causal training shows world and action modeling improve each other — IAAI233 · 2026-10-16
- Baidu's Virtual Head Attention tames the KV cache, FLOPs, quality triangle in long-context LLMs — baidu · 2026-10-16
- 5.8M-param Speech Event Detector adds laughs, breaths and背景 sounds to Whisper transcripts — Ayush Kumar Bar · 2026-10-16