BOHM: Zero-Cost Hierarchical Attribution for Compound AI Systems

Researchers have introduced BOHM, a zero-cost attribution method for compound AI systems that extracts hierarchical attribution trees directly from existing routing weights. The method provides multi-resolution attribution at every level simultaneously without requiring access to component internals, unlike Shapley-based approaches that demand costly coalition evaluations. In tests across 18 LLMs, 5-driver agentic systems, and a US Census hierarchy, BOHM achieved strong ranking correlations with ground truth while operating at a fraction of SHAP's computational cost.

arXiv:2605.22866v1 Announce Type: new Abstract: Compound AI systems route tasks through hierarchies of specialised components. Attribution is dominated by Shapley-based methods SHAP , which decompose a coalition value function into per-component marginal contributions and require evaluation of the system on arbitrary component subsets. That requirement fails for third-party APIs, opaque endpoints, and agentic orchestrators that concentrate routing on a few tools, leaving most coalitions un-evaluable from the deployed orchestrator. We introduce BOHM, which extracts a hierarchical attribution tree directly from the routing weights such systems already maintain: leaf attribution is the path product of root-to-leaf routing weights; level-k attribution is the induced distribution over depth-k nodes. The method has zero marginal cost, requires no access to component internals, and provides multi-resolution attribution at every level simultaneously, which flat methods cannot offer at any evaluation budget. BOHM and SHAP answer different questions and converge when the deployed router routes near-optimally. On 18 LLMs in a 3-level hierarchy over 880 LiveCodeBench problems, BOHM yields Kendall tau=0.928; SHAP reaches tau=0.980 at 9,000x more coalition evaluations per seed. On a 5-driver, 7-benchmark agentic study 35 cells, complete coverage , drivers concentrate routing on a single tool top-share median 0.65 , and cell-level tau BOHM,SHAP is predicted by whether the driver's top pick is the empirically best tool mean +0.22 vs ~+0.01 . On a US Census hierarchy 475 leaves, 4 levels , BOHM recovers ground-truth rankings at every level tau up to 0.722 . BOHM satisfies efficiency, monotonicity, symmetry, and weak suppression but not Shapley's additivity. It is best understood as a complementary primitive: a multi-resolution decomposition computable wherever routing state exists, whose disagreement with Shapley is itself diagnostic.