Preregistered LENR Evidence Gate (PLEG)

This paper develops a revised mathematical solution framework for the chemistry challenge "Establish reproducible experimental evidence for any claimed anomalous low-energy nuclear reaction?". The proposed Preregistered LENR Evidence Gate (PLEG) was constructed after an earlier formulation failed or remained only partially supported. The specific failure was: The earlier framing risked treating anomalous heat alone as nuclear evidence. The revision is: Use a preregistered evidence gate requiring closed calorimetry, nuclear products, negative blanks and independent replication. The resulting central relation is Evidence=I[Qexcess≫σ]·I[nuclear products]·I[blank negative]·I[independent replication]. Unlike a generic logarithmic template, the present model is tied to problem-specific state variables, an explicit inference or decision rule, and a falsification protocol. External validation is based on primary research studies selected for direct relevance to the variables or decision represented by the model. The multi-institution cold-fusion re-evaluation remains negative. The model passes because it correctly returns NULL rather than forcing a positive result. The key quantitative or operational anchor is: No reproducible anomalous nuclear effect satisfying the strict criteria was established. The paper distinguishes model validation from completion of the full grand target: a model may pass by correctly predicting a partial, negative, or unresolved experimental state. We derive the model, analyze identifiability and sensitivity, formulate prospective tests, and state conditions under which the model must be rejected or expanded.