Synapse: Federated Tool Routing via Typed Compendium Artifacts
This work addresses the problem of federated learning with heterogeneous, frozen LLMs where no shared data or weights are allowed, providing a principled alternative to flat units.
The authors propose typed federated artifacts (schema-validated objects) to enable privacy, conflict resolution, and cross-model transfer in federated learning, instantiated as SYNAPSE. Their compendium achieves cross-architectural transfer across four LLM families with approximately 2 pt loss, which weight-sharing federation cannot provide.
The unit of collaboration in federated learning determines what guarantees are even expressible. Flat units like weights, prompts, raw examples, carry no type signature on which privacy, conflict resolution, or cross-model transfer can dispatch as well-defined operations. We propose typed federated artifacts: schema validated objects whose declared field structure makes per field differential privacy, schema aware merging, and cross architectural transfer first-class operations rather than heuristic approximations. We instantiate this as SYNAPSE, a compendium for federated tool routing across clients with frozen, heterogeneous LLMs and no shared data or weights which is a setting flat units cannot handle without either leaking gradients or discarding structure. The compendium admits a typed merge operator with field wise conflict resolution, a formal DP guarantee on numeric metadata, and conditional retrieval distortion and routing-stability results empirically characterized on five distributions, including one where the contraction premise fails. A single compendium transfers across four LLM families (LLaMA 3.18B,LLaMA 3.2-3B, Mistral 7B, GPT 4o) with approximately 2 pt loss, a capability weight-sharing federation cannot provide without architectural matching.