SkillGym trains Qwen3.5-35B-A3B on 2,756 skill environments
Continual training on distilled skills outperforming a larger-context approach suggests skill libraries can be baked into weights rather than retrieved at inference time. Teams building terminal or coding agents can test whether the same recipe transfers to their own models and checkers. The result also matters for anyone benchmarking open-weight Qwen models against Claude Sonnet 4.6 and GPT-5.4 Mini on agentic tasks.
A paper proposes SkillGym, a method that converts human-written skills into 2,756 training environments with code-based checkers.
The authors fine-tune Qwen3.5-35B-A3B on 8,364 successful trajectories generated in these environments.
After training, the model running in Claude Code gains 19.10 points on Terminal-Bench 2.
1 and 28.13 points on skill-assisted SkillsBench v1.
1, reaching 51.47%.
On SkillsBench v1.1 this score is above the reported results for Claude Sonnet 4.
6 and GPT-5.4 Mini.
Without skills loaded, the trained model still beats its base model using skills in context.
An interesting idea is to continually train specialized models on skills.
This paper explores that idea.
They propose SkillGym which turns human-written skills into 2,756 training environments with code-based checkers, then fine-tunes on 8,364 successful trajectories.
After fine-tuning, Qwen3.5-35B-A3B running in Claude Code gains 19.10 points on Terminal-Bench 2.1 and 28.13 points on skill-assisted SkillsBench v1.1, where it reaches 51.47%.
That is above the reported scores for Claude Sonnet 4.6 and GPT-5.4 Mini.
With no skills loaded, the trained model beats the base model that has the skills in context.
Paper: https://t.co/ubb53Qd1Ms
This report is categorized under AITOP's "Papers" desk, focusing on China's AI industry and domestic foundation model ecosystem. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Qwen, SkillGym, fine-tuning, agents.
Source attribution: curated from DAIR.AI (@dair_ai). Cross-source corroboration: 3 independent sources have covered this development (including AIGCLINK, Gorden Sun), confirming industry-wide signal strength.
