arXiv subjects

Mahdi Eslamimehr

Publications and source records attributed to Mahdi Eslamimehr.

6 recordsLinked to original sources

arXiv2026

Directed Neuro-Symbolic Stochastic Execution for Verification of Distributed Parallel AI Programs

Distributed parallel Artificial Intelligence (AI) programs expose reliability gaps that conventional testing cannot close: parallel executions are non-deterministic, and AI workloads bring high-dimensional inputs and non-linear operations that defeat fuzzing and symbolic execution in isolation. We present Directed Neuro-Symbolic Stochastic Execution (DNSSE), a hybrid testing framework that couples schedule prediction guided by a Large Language Model (LLM) with symbolic constraint solving and coverage-guided stochastic mutation. We model distributed AI executions as non-deterministic transition systems, specify correctness in linear temporal logic, and prove soundness, bounded completeness, and probabilistic completeness of the hybrid solver, together with an expected-cost analysis of LLM-guided schedule exploration. A scalable implementation on PyTorch and Ray detects 2.9% more concurrency bugs than the strongest baseline and raises average branch coverage from 68.6 % to 91.6 % across five realistic distributed AI benchmarks.

cs.AI↗

arXiv2026

SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI

High-impact generative AI makes catastrophic misuse a lifecycle-control problem, not merely a prompt-filtering problem. SAGE is a safety-first, authorization-separated architecture in which credible catastrophic-enablement risk constrains admissibility before utility, latency, or commercial objectives are considered. It combines signed release manifests, diverse detectors, robust risk envelopes, least-risk defaults, output checking, three-valued monitoring, protected audit chains, containment, and rollback. Formal results establish safety priority, conservative detector bounds, monotone release gating, tamper-evident records, and an authorization cut; two PRISM abstractions verify authorization separation and lifecycle invariants under explicit assumptions. A frozen, vendor-symmetric study sent 84 cases to each of four GPT, four Claude, and two Gemini snapshots: 840 calls yielded 794 target responses, 46 provider errors, and 449 successful judgments covering 375 responses. Eight snapshots had complete judged domain coverage. Harmful-compliance estimates were low; variation arose mainly from benign utility and safe redirection. Seven multiplicity-adjusted contrasts involving Claude, Gemini, or GPT-5 snapshots and the GPT-5 mini and GPT-5 nano snapshots were supported, while no tested contrast between the Claude or Gemini snapshots and GPT-5 or GPT-5.5 survived correction. The observed harmful-compliance range is a conservative, protocol-bound view from one generation per prompt with no tools, retrieval, history, or human adjudication; it is not an upper bound on operational assistance. A preregistered extension specifies how to test a wider best-worst gap using a locked split, repeated sampling, multi-turn and sandboxed-tool conditions, and domain-expert scoring.

cs.AI↗

arXiv2026

Synergistic Directed Execution and LLM-Driven Analysis for Zero-Day AI-Generated Malware Detection

The weaponization of LLMs for automated malware generation poses an existential threat to conventional detection paradigms. AI-generated malware exhibits polymorphic, metamorphic, and context-aware evasion capabilities that render signature-based and shallow heuristic defenses obsolete. This paper introduces a novel hybrid analysis framework that synergistically combines \emph{concolic execution} with \emph{LLM-augmented path prioritization} and \emph{deep-learning-based vulnerability classification} to detect zero-day AI-generated malware with provable guarantees. We formalize the detection problem within a first-order temporal logic over program execution traces, define a lattice-theoretic abstraction for path constraint spaces, and prove both the \emph{soundness} and \emph{relative completeness} of our detection algorithm, assuming classifier correctness. The framework introduces three novel algorithms: (i) an LLM-guided concolic exploration strategy that reduces the average number of explored paths by 73.2\% compared to depth-first search while maintaining equivalent malicious-path coverage; (ii) a transformer-based path-constraint classifier trained on symbolic execution traces; and (iii) a feedback loop that iteratively refines the LLM's prioritization policy using reinforcement learning from detection outcomes. We provide a comprehensive implementation built upon \texttt{angr} 9.2, \texttt{Z3} 4.12, Hugging Face Transformers 4.38, and PyTorch 2.2, with configuration details enabling reproducibility. Experimental evaluation on the EMBER, Malimg, SOREL-20M, and a novel AI-Gen-Malware benchmark comprising 2{,}500 LLM-synthesized samples demonstrates that achieves 98.7\% accuracy on conventional malware and 97.5\% accuracy on AI-generated threats, outperforming ClamAV, YARA, MalConv, and EMBER-GBDT baselines by margins of 8.4--52.2 percentage points on AI-generated samples.

cs.CR↗

arXiv2026

MOTION: ML-Assisted On-Device Low-Latency Motion Recognition

The use of tiny devices capable of low-latency gesture recognition is gaining momentum in everyday human-computer interaction and especially in medical monitoring fields. Embedded solutions such as fall detection, rehabilitation tracking, and patient supervision require fast and efficient tracking of movements while avoiding unwanted false alarms. This study presents an efficient solution on how to build very efficient motion-based models only using triaxial accelerometer sensors. We explore the capability of the AutoML pipelines to extract the most important features from the data segments. This approach also involves training multiple lightweight machine learning algorithms using the extracted features. We use WeBe Band, a multi-sensor wearable device that is equipped with a powerful enough MCU to effectively perform gesture recognition entirely on the device. Of the models explored, we found that the neural network provided the best balance between accuracy, latency, and memory use. Our results also demonstrate that reliable real-time gesture recognition can be achieved in WeBe Band, with great potential for real-time medical monitoring solutions that require a secure and fast response time.

cs.CV↗

arXiv2026

Hybrid Concolic Testing with Large Language Models for Guided Path Exploration

Concolic testing, a powerful hybrid software testing technique, has historically been plagued by fundamental limitations such as path explosion and the high cost of constraint solving, which hinder its practical application in large-scale, real-world software systems. This paper introduces a novel algorithmic framework that synergistically integrates concolic execution with Large Language Models (LLMs) to overcome these challenges. Our hybrid approach leverages the semantic reasoning capabilities of LLMs to guide path exploration, prioritize interesting execution paths, and assist in constraint solving. We formally define the system architecture and algorithms that constitute this new paradigm. Through a series of experiments on both synthetic and real-world Fintech applications, we demonstrate that our approach significantly outperforms traditional concolic testing, random testing, and genetic algorithm-based methods in terms of branch coverage, path coverage, and time-to-coverage. The results indicate that by combining the strengths of both concolic execution and LLMs, our method achieves a more efficient and effective exploration of the program state space, leading to improved bug detection capabilities.

cs.SE↗

arXiv2018

Recognising and Generating Terms using Derivatives of Parsing Expression Grammars

Grammar-based sentence generation has been thoroughly explored for Context-Free Grammars (CFGs), but remains unsolved for recognition-based approaches such as Parsing Expression Grammars (PEGs). Lacking tool support, language designers using PEGs have difficulty predicting the behaviour of their parsers. In this paper, we extend the idea of derivatives, originally formulated for regular expressions, to PEGs. We then present a novel technique for sentence generation based on derivatives, applicable to any grammatical formalism for which the derivative can be defined--now including PEGs. Finally, we propose applying derivatives more generally to other problems facing language designers and implementers.

cs.PL↗