Search papers, labs, and topics across Lattice.
This paper introduces the Scientific Explanation-Admissibility Machines (SEAM), a framework designed to assess the global consistency of explanations generated by scientific machine learning models, addressing the limitations of local validation methods. SEAM-$\Omega$ utilizes a structured explanation representation to compare neighboring explanations, identify inconsistencies, and test competing accounts by restricting repairs to permissible revisions. Through extensive experimentation, SEAM demonstrates its capability to detect incompatible explanations even when local predictions are accurate, thereby enhancing the reliability of scientific models in diverse contexts.
SEAM reveals that even accurate local predictions can mask significant global inconsistencies in scientific explanations, challenging traditional validation approaches.
Scientific machine learning commonly validates models at the level of a subdomain, a benchmark split, or an explanation for one prediction. Yet such local checks cannot establish whether the resulting explanations can be assembled into one globally admissible explanation. We introduce Scientific Explanation-Admissibility Machines (SEAM), a generator-agnostic framework that makes this local-to-global consistency question computable across regions, sensors, regimes, and model components. The finite explanation-sheaf instantiation SEAM-$\Omega$ represents each region by a structured explanation with state, closure, and observation channels together with optional contract metadata; compares neighboring explanations on their overlaps; and converts disagreement into a channel-resolved obstruction. This obstruction locates inconsistency and tests competing declared accounts by restricting each repair to the revisions that one account permits. Exact feasibility refutes or retains an account; when exact repair is unavailable, residual-aware regularized records provide a separately labeled empirical attribution. The framework also separates inconsistency from non-identifiability and monitors learned generators under distribution shift. We establish theorems for minimum-cost intervention and conservation-contract detectability, together with companion results for identifiability and closure recoverability. Across nineteen experiments involving synthetic partial differential equation systems and out-of-distribution Fourier neural operator (FNO) monitoring, SEAM detects incompatible explanations even when local predictions are accurate, and attributes failures to specific channels and overlaps. SEAM adds a global explanation-consistency audit to existing solvers and learning models, testing whether their local explanations form a coherent scientific account.
