Tsinghua + Qwen paper: rebuilding agent workspaces beats imitating trajectories, lifts Terminal-Bench to 58.1%
rohanpaul_ai · x · 2026-09-10
A new Tsinghua + Qwen team paper flips the usual agent-data recipe: instead of imitating recorded trajectories, reconstruct the actual code workspace so you can inject new bugs, new features, or stronger agents and generate many fresh verified training tasks.
- Trajectory = one frozen solution; environment = a reusable codebase producing many tasks.
- Training Qwen3.5-27B on this data moved Terminal-Bench 2.1 from 46.2% to 58.1% and EvoCode-Bench v2 MT@4 from 6.3 to 20.1 (Terminus2 setup).
- Takeaway: collect reusable execution environments, not more traces to imitate.
More from coding & agent
- AWS on agent prompt caching: explicit + implicit patterns to slash token costs — AWS ML Blog · 2026-10-27
- He gave his AI agent a phone number and email — then hired it as social media manager — heyshrutimishra · 2026-10-27
- Firecrawl launches Alexandria company-data API and announces $75M Series B — devdigest · 2026-10-27
- 'Let an agent run wild on vague goals and it makes your code worse 9 out of 10 times' — eptwts · 2026-10-27
- LangChain to demo the in-house design agent its team uses daily, this Wednesday — LangChain · 2026-10-27
- Biotech Agent Series Day 8: A Notebook Review Agent to Auto-Audit Lab Records — nlarusstone · 2026-10-27