Two linear maps
A frozen Qwen3-32B routed skills better than pipelines that bolted on up to 16B parameters
Tsinghua researchers report that two trained linear maps, read from a frozen Qwen3-32B, route skills better than routers that add parameters.
Gavel projects the task and each skill through the maps, then resumes the shortlisted skills’ forward passes for a yes-or-no verdict. It scores up to 13.4 points above progressive-disclosure and retrieve-and-rerank pipelines that bolt on 1.2B to 16B parameters. When the need for a skill arises mid-rollout, the margin is 21.9 points above the same pipelines. No skill text sits in the context at all.
What it changes If your harness preloads every skill’s metadata, this is the argument for stopping; the code is the authors’ and the backbone is Qwen3-32B.
Primary source read
The Router Within: Eliciting Native Skill Routing from a Frozen LLM