Work

The Pre-Flight Check for Autonomous AI: Zero-Model Structural Reasoning Validation at Scale

Year: 2026

Type: report

Abstract: We present the Subtractive Filter, a lightweight, model-free reasoning integrity validator for Large Language Model (LLM) outputs. Operating on deterministic pattern matching, it requires zero model inference and achieves 82,544 samples/second throughput. Evaluated on 58,293 samples from the HaluEval benchmark, the filter demonstrates 91.3% F1 on structural reasoning failures (contradictions, circular logic) while scoring 4.0% F1 on factual hallucinations. This paper argues for pre-execution structural reasoning validation as a critical, distinct layer in AI safety, analogous to aviation pre-flight checks.

DOI: 10.5281/zenodo.18672141

Author: Abdessattar, Moez

Language: English

Topic: Adversarial Robustness in Machine Learning

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: gold

Funder: Nederlandse Organisatie voor Wetenschappelijk Onderzoek

Award: Modeling semantic contextual computations in the human brain: algorithms, neuromodulation, and cognitive function

OpenAlex ID: W7129304036

Cites:

Cited by:

Related to: