Search papers, labs, and topics across Lattice.
EvolveNet introduces a novel approach to collaborative harness evolution for LLM agents, allowing independent adaptations of a shared harness based on localized experience rather than centralized optimization. This method addresses the limitations of traditional single-optimizer frameworks by enabling concurrent evolutionary searches across diverse agent deployments, thereby enhancing the harness's capabilities without requiring model weight updates. The results demonstrate significant improvements across various tasks, particularly in heterogeneous environments, highlighting the effectiveness of compositional learning from diverse agent experiences.
EvolveNet reveals that decentralized evolution of agent harnesses can lead to substantial performance gains by leveraging localized experience rather than relying on centralized optimization.
The capabilities of an LLM agent depend not only on its model but on the harness: the executable program that constructs context, invokes tools, verifies results, and recovers from failure. Recent work shows that evolving the harness yields persistent improvements without updating model weights. Existing approaches, however, assume that all execution experience can be routed to a single optimizer, which evolves one harness along a sequential trajectory. Real agent ecosystems violate that assumption: users, organizations, and environments generate isolated streams of experience that cannot be pooled, so the experience most worth learning from is exactly the experience that cannot be directly centralized. We introduce EvolveNet, a paradigm of collaborative harness evolution that moves experience extraction to the data. A shared harness is broadcast to data-local agent deployments, each of which evolves it on its own workload. Only the resulting program adaptations are composed into an updated shared harness and redistributed, so that every participating agent inherits operational experience discovered by the others. By shifting the aggregation boundary from raw workloads to learned adaptations, EvolveNet keeps workloads local and allows multiple evolutionary searches to proceed concurrently with reduced serial depth. Because independently modified programs cannot be averaged like model parameters and may conflict when composed, EvolveNet introduces scope-typed, evidence-guided program aggregation. Across five settings spanning text-to-SQL, data-science coding, competitive programming, software engineering, and agentic workflows, EvolveNet improves the shared harness in all five, with the largest gains under heterogeneous workloads, and ablations attribute the improvement to composition of adaptations from different agents rather than to selecting among them.
