arXiv subjects

Gustavo Quiros Araya

Publications and source records attributed to Gustavo Quiros Araya.

6 recordsLinked to original sources

arXiv2025

LLM4CVE: Enabling Iterative Automated Vulnerability Repair with Large Language Models

Software vulnerabilities continue to be ubiquitous, even in the era of AI-powered code assistants, advanced static analysis tools, and the adoption of extensive testing frameworks. It has become apparent that we must not simply prevent these bugs, but also eliminate them in a quick, efficient manner. Yet, human code intervention is slow, costly, and can often lead to further security vulnerabilities, especially in legacy codebases. The advent of highly advanced Large Language Models (LLM) has opened up the possibility for many software defects to be patched automatically. We propose LLM4CVE an LLM-based iterative pipeline that robustly fixes vulnerable functions in real-world code with high accuracy. We examine our pipeline with State-of-the-Art LLMs, such as GPT-3.5, GPT-4o, Llama 38B, and Llama 3 70B. We achieve a human-verified quality score of 8.51/10 and an increase in groundtruth code similarity of 20% with Llama 3 70B. To promote further research in the area of LLM-based vulnerability repair, we publish our testing apparatus, fine-tuned weights, and experimental data on our website

cs.SE↗

arXiv2024

Behavior Trees with Dataflow: Coordinating Reactive Tasks in Lingua Franca

Behavior Trees (BTs) provide a lean set of control flow elements that are easily composable in a modular tree structure. They are well established for modeling the high-level behavior of non-player characters in computer games and recently gained popularity in other areas such as industrial automation. While BTs nicely express control, data handling aspects so far must be provided separately, e. g. in the form of blackboards. This may hamper reusability and can be a source of nondeterminism. We here present a dataflow extension to BTs that explicitly models data relations and communication. We provide a combined textual/graphical approach in line with modern, productivity-enhancing pragmatics-aware modeling techniques. We realized and validated that approach in the recently introduced polyglot coordination language Lingua Franca (LF).

cs.PL↗

arXiv2024

LLM4PLC: Harnessing Large Language Models for Verifiable Programming of PLCs in Industrial Control Systems

Although Large Language Models (LLMs) have established pre-dominance in automated code generation, they are not devoid of shortcomings. The pertinent issues primarily relate to the absence of execution guarantees for generated code, a lack of explainability, and suboptimal support for essential but niche programming languages. State-of-the-art LLMs such as GPT-4 and LLaMa2 fail to produce valid programs for Industrial Control Systems (ICS) operated by Programmable Logic Controllers (PLCs). We propose LLM4PLC, a user-guided iterative pipeline leveraging user feedback and external verification tools including grammar checkers, compilers and SMV verifiers to guide the LLM's generation. We further enhance the generation potential of LLM by employing Prompt Engineering and model fine-tuning through the creation and usage of LoRAs. We validate this system using a FischerTechnik Manufacturing TestBed (MFTB), illustrating how LLMs can evolve from generating structurally flawed code to producing verifiably correct programs for industrial applications. We run a complete test suite on GPT-3.5, GPT-4, Code Llama-7B, a fine-tuned Code Llama-7B model, Code Llama-34B, and a fine-tuned Code Llama-34B model. The proposed pipeline improved the generation success rate from 47% to 72%, and the Survey-of-Experts code quality from 2.25/10 to 7.75/10. To promote open research, we share the complete experimental setup, the LLM Fine-Tuning Weights, and the video demonstrations of the different programs on our dedicated webpage.

cs.SE↗

arXiv2023

Resiliency Analysis of LLM generated models for Industrial Automation

This paper proposes a study of the resilience and efficiency of automatically generated industrial automation and control systems using Large Language Models (LLMs). The approach involves modeling the system using percolation theory to estimate its resilience and formulating the design problem as an optimization problem subject to constraints. Techniques from stochastic optimization and regret analysis are used to find a near-optimal solution with provable regret bounds. The study aims to provide insights into the effectiveness and reliability of automatically generated systems in industrial automation and control, and to identify potential areas for improvement in their design and implementation.

cs.SE↗

arXiv2023

No Code AI: Automatic generation of Function Block Diagrams from documentation and associated heuristic for context-aware ML algorithm training

Industrial process engineering and PLC program development have traditionally favored Function Block Diagram (FBD) programming over classical imperative style programming like the object oriented and functional programming paradigms. The increasing momentum in the adoption and trial of ideas now classified as 'No Code' or 'Low Code' alongside the mainstream success of statistical learning theory or the so-called machine learning is redefining the way in which we structure programs for the digital machine to execute. A principal focus of 'No Code' is deriving executable programs directly from a set of requirement documents or any other documentation that defines consumer or customer expectation. We present a method for generating Function Block Diagram (FBD) programs as either the intermediate or final artifact that can be executed by a target system from a set of requirement documents using a constrained selection algorithm that draws from the top line of an associated recommender system. The results presented demonstrate that this type of No-code generative model is a viable option for industrial process design.

cs.SE↗

arXiv2023

Comparative Analysis of CHATGPT and the evolution of language models

Interest in Large Language Models (LLMs) has increased drastically since the emergence of ChatGPT and the outstanding positive societal response to the ease with which it performs tasks in Natural Language Processing (NLP). The triumph of ChatGPT, however, is how it seamlessly bridges the divide between language generation and knowledge models. In some cases, it provides anecdotal evidence of a framework for replicating human intuition over a knowledge domain. This paper highlights the prevailing ideas in NLP, including machine translation, machine summarization, question-answering, and language generation, and compares the performance of ChatGPT with the major algorithms in each of these categories using the Spontaneous Quality (SQ) score. A strategy for validating the arguments and results of ChatGPT is presented summarily as an example of safe, large-scale adoption of LLMs.

cs.CL↗