CHRONOS: Multi-Agent Coordination for Evolving Data Marketplaces
A new architecture called CHRONOS addresses three failures in temporal knowledge-graph data marketplaces: stale hybrid index shortcuts, stationary Shapley pricing, and uncoordinated agents over-consuming differential-privacy budgets. The three-layer system uses neural-ODE temporal decay, changepoint-conditioned Shapley valuation, and EXP3-IX for regret minimization while enforcing differential privacy. The paper is available on arXiv under identifier 2605.23887.
Key facts
- CHRONOS is a three-layer architecture for temporal knowledge-graph data marketplaces.
- It addresses stale hybrid index shortcuts, stationary Shapley pricing, and uncoordinated agent behavior.
- Layer one uses neural-ODE temporal decay to provide a per-query expected recall-loss bound.
- Layer two conditions Shapley valuation on detected changepoints.
- Layer three uses EXP3-IX to achieve regret of O(√T log T) while enforcing differential privacy.
- The paper is published on arXiv with identifier 2605.23887.
- The architecture provides explicit public and private separation.
- The monotone-envelope guarantee reduces bound looseness to 1.8–3.2 times observed loss.
Entities
Institutions
- arXiv