Queen's University Belfast

Biomedical retrieval sentence encoders can be converted into typed decision models that return thresholdable probabilities in one forward pass, with a cross head reaching 66–73 seen-task points, close to fine-tuned Laya-large at 75.0 with about three times the parameters.

Biomedical retrieval sentence encoders can be converted into typed decision models that return thresholdable probabilities in one forward pass, with a cross head reaching 66–73 seen-task points, close to fine-tuned Laya-large at 75.0 with about three times the parameters.

This paper develops a formal framework that records, for each workflow unit, its requirement, approved operational formalization, participation and assurance regime, mechanism, fallback, evidence obligations and readiness status, distinguishing deterministic verification, statistically calibrated admission, authorized human judgement or adjudication, retained deterministic tool paths and explicit non-participation; the authors execute it on a 17-unit wing-spar structural-analysis workflow where an AI-generated CAD program passed all 23 deterministic checks at the first attempt, while stress clauses did not yield PASS because the predeclared mesh-convergence evidence was insufficient and were referred to engineering adjudication.

This paper develops a formal framework that records, for each workflow unit, its requirement, approved operational formalization, participation and assurance regime, mechanism, fallback, evidence obligations and readiness status, distinguishing deterministic verification, statistically calibrated admission, authorized human judgement or adjudication, retained deterministic tool paths and explicit non-participation; the authors execute it on a 17-unit wing-spar structural-analysis workflow where an AI-generated CAD program passed all 23 deterministic checks at the first attempt, while stress clauses did not yield PASS because the predeclared mesh-convergence evidence was insufficient and were referred to engineering adjudication.

Addressing the lack of a systematic survey and in-depth analysis of the latest methodologies in high-level synthesis for approximate computing (AHLS), this survey summarizes recent technologies in the field with particular focus on error estimation, approximation techniques, and design space exploration (DSE), and analyzes current research gaps, aiming to give researchers, engineers, and scholars a theoretical and practical framework for AHLS.

Addressing the lack of a systematic survey and in-depth analysis of the latest methodologies in high-level synthesis for approximate computing (AHLS), this survey summarizes recent technologies in the field with particular focus on error estimation, approximation techniques, and design space exploration (DSE), and analyzes current research gaps, aiming to give researchers, engineers, and scholars a theoretical and practical framework for AHLS.