arXiv subjects

Thomas Runkler

Publications and source records attributed to Thomas Runkler.

At least 19 recordsLinked to original sources

arXiv2026

Agent4cs: A Multi-agent System for Code Summarization in Large Hierarchical Codebases

Understanding large, complex codebases, especially those with obfuscated structures and incomplete documentation, remains a significant challenge. Existing code summarization solutions often rely on a single language model or coding assistant like Claude Code, and treat source code as flat text, underutilizing the rich interdependencies and hierarchical information within a repository. To address these shortcomings, we propose Agent4cs - a multi-agent framework that summarizes large codebases in a bottom-up fashion, where a summarization agent focuses on producing robust summaries; a keyword-extraction agent proactively identifies critical information from subfolders; and a quality-assurance agent iteratively refines the outputs for readability, coherence, and completeness. Evaluated on 7 frontier models, Agent4cs improves semantic consistency across all folder levels by average 8% compared to two structured prompting baselines with code segments. Furthermore, extensive evaluation on real-world datasets demonstrates up to 38% gains in normalized keyword coverage rate over the same baselines.

cs.AI↗

arXiv2026

LLM-Based Agentic Systems for Software Engineering: Challenges and Opportunities

Despite recent advancements in Large Language Models (LLMs), complex Software Engineering (SE) tasks require more collaborative and specialized approaches. This concept paper systematically reviews the emerging paradigm of LLM-based multi-agent systems, examining their applications across the Software Development Life Cycle (SDLC), from requirements engineering and code generation to static code checking, testing, and debugging. We delve into a wide range of topics such as language model selection, SE evaluation benchmarks, state-of-the-art agentic frameworks and communication protocols. Furthermore, we identify key challenges and outline future research opportunities, with a focus on multi-agent orchestration, human-agent coordination, computational cost optimization, and effective data collection. This work aims to provide researchers and practitioners with valuable insights into the current forefront landscape of agentic systems within the software engineering domain.

cs.SE↗

arXiv2025

The Few-shot Dilemma: Over-prompting Large Language Models

Over-prompting, a phenomenon where excessive examples in prompts lead to diminished performance in Large Language Models (LLMs), challenges the conventional wisdom about in-context few-shot learning. To investigate this few-shot dilemma, we outline a prompting framework that leverages three standard few-shot selection methods - random sampling, semantic embedding, and TF-IDF vectors - and evaluate these methods across multiple LLMs, including GPT-4o, GPT-3.5-turbo, DeepSeek-V3, Gemma-3, LLaMA-3.1, LLaMA-3.2, and Mistral. Our experimental results reveal that incorporating excessive domain-specific examples into prompts can paradoxically degrade performance in certain LLMs, which contradicts the prior empirical conclusion that more relevant few-shot examples universally benefit LLMs. Given the trend of LLM-assisted software engineering and requirement analysis, we experiment with two real-world software requirement classification datasets. By gradually increasing the number of TF-IDF-selected and stratified few-shot examples, we identify their optimal quantity for each LLM. This combined approach achieves superior performance with fewer examples, avoiding the over-prompting problem, thus surpassing the state-of-the-art by 1% in classifying functional and non-functional requirements.

cs.CL↗

arXiv2025

Is Q-learning an Ill-posed Problem?

This paper investigates the instability of Q-learning in continuous environments, a challenge frequently encountered by practitioners. Traditionally, this instability is attributed to bootstrapping and regression model errors. Using a representative reinforcement learning benchmark, we systematically examine the effects of bootstrapping and model inaccuracies by incrementally eliminating these potential error sources. Our findings reveal that even in relatively simple benchmarks, the fundamental task of Q-learning - iteratively learning a Q-function from policy-specific target values - can be inherently ill-posed and prone to failure. These insights cast doubt on the reliability of Q-learning as a universal solution for reinforcement learning problems.

cs.LG↗

arXiv2024

Conceptual In-Context Learning and Chain of Concepts: Solving Complex Conceptual Problems Using Large Language Models

Science and engineering problems fall in the category of complex conceptual problems that require specific conceptual information (CI) like math/logic -related know-how, process information, or engineering guidelines to solve them. Large Language Models (LLMs) are promising agents to solve such complex conceptual problems due to their implications in advancing engineering and science tasks like assisted problem-solving. But vanilla LLMs, trained on open-world data, lack the necessary CI. In this work, we specifically explore shallow customization methods (SCMs) of LLMs for solving complex conceptual problems. We propose two novel SCM algorithms for LLM, to augment LLMs with CI and enable LLMs to solve complex conceptual problems: Conceptual In-Context Learning (C-ICL) and Chain of Concepts (CoC). The problem tackled in this paper is generation of proprietary data models in the engineering/industry domain based on conceptual information in data modelling guidelines. We evaluate our algorithms on varied sizes of the OpenAI LLMs against four evaluation metrics related to syntactic and semantic correctness, time and cost incurred. The proposed algorithms perform better than currently popular LLM SCMs like In-context Learning (ICL) and Chain of Thoughts (CoT). It was observed that as compared to CoT, response correctness increased by 30.6% and 29.88% for the new SCMs C-ICL and CoC respectively. Qualitative analysis suggests that the proposed new SCMs activate emergent capabilities in LLMs, previously unobserved in the existing SCMs. They make problem-solving processes more transparent and reduce hallucinations and the tendency of model responses to copy examples from prompts (parroting).

cs.CL↗

arXiv2024

FsPONER: Few-shot Prompt Optimization for Named Entity Recognition in Domain-specific Scenarios

Large Language Models (LLMs) have provided a new pathway for Named Entity Recognition (NER) tasks. Compared with fine-tuning, LLM-powered prompting methods avoid the need for training, conserve substantial computational resources, and rely on minimal annotated data. Previous studies have achieved comparable performance to fully supervised BERT-based fine-tuning approaches on general NER benchmarks. However, none of the previous approaches has investigated the efficiency of LLM-based few-shot learning in domain-specific scenarios. To address this gap, we introduce FsPONER, a novel approach for optimizing few-shot prompts, and evaluate its performance on domain-specific NER datasets, with a focus on industrial manufacturing and maintenance, while using multiple LLMs -- GPT-4-32K, GPT-3.5-Turbo, LLaMA 2-chat, and Vicuna. FsPONER consists of three few-shot selection methods based on random sampling, TF-IDF vectors, and a combination of both. We compare these methods with a general-purpose GPT-NER method as the number of few-shot examples increases and evaluate their optimal NER performance against fine-tuned BERT and LLaMA 2-chat. In the considered real-world scenarios with data scarcity, FsPONER with TF-IDF surpasses fine-tuned models by approximately 10% in F1 score.

cs.CL↗

arXiv2024

Fusion of Domain-Adapted Vision and Language Models for Medical Visual Question Answering

Vision-language models, while effective in general domains and showing strong performance in diverse multi-modal applications like visual question-answering (VQA), struggle to maintain the same level of effectiveness in more specialized domains, e.g., medical. We propose a medical vision-language model that integrates large vision and language models adapted for the medical domain. This model goes through three stages of parameter-efficient training using three separate biomedical and radiology multi-modal visual and text datasets. The proposed model achieves state-of-the-art performance on the SLAKE 1.0 medical VQA (MedVQA) dataset with an overall accuracy of 87.5% and demonstrates strong performance on another MedVQA dataset, VQA-RAD, achieving an overall accuracy of 73.2%.

cs.CL↗

arXiv2024

Wiki-TabNER: Integrating Named Entity Recognition into Wikipedia Tables

Interest in solving table interpretation tasks has grown over the years, yet it still relies on existing datasets that may be overly simplified. This is potentially reducing the effectiveness of the dataset for thorough evaluation and failing to accurately represent tables as they appear in the real-world. To enrich the existing benchmark datasets, we extract and annotate a new, more challenging dataset. The proposed Wiki-TabNER dataset features complex tables containing several entities per cell, with named entities labeled using DBpedia classes. This dataset is specifically designed to address named entity recognition (NER) task within tables, but it can also be used as a more challenging dataset for evaluating the entity linking task. In this paper we describe the distinguishing features of the Wiki-TabNER dataset and the labeling process. In addition, we propose a prompting framework for evaluating the new large language models on the within tables NER task. Finally, we perform qualitative analysis to gain insights into the challenges encountered by the models and to understand the limitations of the proposed~dataset.

cs.AI↗

arXiv2023

Automatic Trade-off Adaptation in Offline RL

Recently, offline RL algorithms have been proposed that remain adaptive at runtime. For example, the LION algorithm \cite{lion} provides the user with an interface to set the trade-off between behavior cloning and optimality w.r.t. the estimated return at runtime. Experts can then use this interface to adapt the policy behavior according to their preferences and find a good trade-off between conservatism and performance optimization. Since expert time is precious, we extend the methodology with an autopilot that automatically finds the correct parameterization of the trade-off, yielding a new algorithm which we term AutoLION.

cs.LG↗

arXiv2022

Detection, Explanation and Filtering of Cyber Attacks Combining Symbolic and Sub-Symbolic Methods

Machine learning (ML) on graph-structured data has recently received deepened interest in the context of intrusion detection in the cybersecurity domain. Due to the increasing amounts of data generated by monitoring tools as well as more and more sophisticated attacks, these ML methods are gaining traction. Knowledge graphs and their corresponding learning techniques such as Graph Neural Networks (GNNs) with their ability to seamlessly integrate data from multiple domains using human-understandable vocabularies, are finding application in the cybersecurity domain. However, similar to other connectionist models, GNNs are lacking transparency in their decision making. This is especially important as there tend to be a high number of false positive alerts in the cybersecurity domain, such that triage needs to be done by domain experts, requiring a lot of man power. Therefore, we are addressing Explainable AI (XAI) for GNNs to enhance trust management by exploring combining symbolic and sub-symbolic methods in the area of cybersecurity that incorporate domain knowledge. We experimented with this approach by generating explanations in an industrial demonstrator system. The proposed method is shown to produce intuitive explanations for alerts for a diverse range of scenarios. Not only do the explanations provide deeper insights into the alerts, but they also lead to a reduction of false positive alerts by 66% and by 93% when including the fidelity metric.

cs.CR↗

arXiv2022

A Metaheuristic Approach for Mining Gradual Patterns

Swarm intelligence is a discipline that studies the collective behavior that is produced by local interactions of a group of individuals with each other and with their environment. In Computer Science domain, numerous swarm intelligence techniques are applied to optimization problems that seek to efficiently find best solutions within a search space. Gradual pattern mining is another Computer Science field that could benefit from the efficiency of swarm based optimization techniques in the task of finding gradual patterns from a huge search space. A gradual pattern is a rule-based correlation that describes the gradual relationship among the attributes of a data set. For example, given attributes {G,H} of a data set a gradual pattern may take the form: "the less G, the more H". In this paper, we propose a numeric encoding for gradual pattern candidates that we use to define an effective search space. In addition, we present a systematic study of several meta-heuristic optimization techniques as efficient solutions to the problem of finding gradual patterns using our search space.

cs.NE↗

arXiv2022

Ant Colony Optimization for Mining Gradual Patterns

Gradual pattern extraction is a field in (KDD) Knowledge Discovery in Databases that maps correlations between attributes of a data set as gradual dependencies. A gradual dependency may take a form of "the more Attribute K , the less Attribute L". In this paper, we propose an ant colony optimization technique that uses a probabilistic approach to learn and extract frequent gradual patterns. Through computational experiments on real-world data sets, we compared the performance of our ant-based algorithm to an existing gradual item set extraction algorithm and we found out that our algorithm outperforms the later especially when dealing with large data sets.

cs.DB↗

arXiv2022

Neuro-symbolic computing with spiking neural networks

Knowledge graphs are an expressive and widely used data structure due to their ability to integrate data from different domains in a sensible and machine-readable way. Thus, they can be used to model a variety of systems such as molecules and social networks. However, it still remains an open question how symbolic reasoning could be realized in spiking systems and, therefore, how spiking neural networks could be applied to such graph data. Here, we extend previous work on spike-based graph algorithms by demonstrating how symbolic and multi-relational information can be encoded using spiking neurons, allowing reasoning over symbolic structures like knowledge graphs with spiking neural networks. The introduced framework is enabled by combining the graph embedding paradigm and the recent progress in training spiking neural networks using error backpropagation. The presented methods are applicable to a variety of spiking neuron models and can be trained end-to-end in combination with other differentiable network architectures, which we demonstrate by implementing a spiking relational graph neural network.

cs.NE↗

arXiv2022

User-Interactive Offline Reinforcement Learning

Offline reinforcement learning algorithms still lack trust in practice due to the risk that the learned policy performs worse than the original policy that generated the dataset or behaves in an unexpected way that is unfamiliar to the user. At the same time, offline RL algorithms are not able to tune their most important hyperparameter - the proximity of the learned policy to the original policy. We propose an algorithm that allows the user to tune this hyperparameter at runtime, thereby addressing both of the above mentioned issues simultaneously. This allows users to start with the original behavior and grant successively greater deviation, as well as stopping at any time when the policy deteriorates or the behavior is too far from the familiar one.

cs.LG↗

arXiv2022

How to Manage Tiny Machine Learning at Scale: An Industrial Perspective

Tiny machine learning (TinyML) has gained widespread popularity where machine learning (ML) is democratized on ubiquitous microcontrollers, processing sensor data everywhere in real-time. To manage TinyML in the industry, where mass deployment happens, we consider the hardware and software constraints, ranging from available onboard sensors and memory size to ML-model architectures and runtime platforms. However, Internet of Things (IoT) devices are typically tailored to specific tasks and are subject to heterogeneity and limited resources. Moreover, TinyML models have been developed with different structures and are often distributed without a clear understanding of their working principles, leading to a fragmented ecosystem. Considering these challenges, we propose a framework using Semantic Web technologies to enable the joint management of TinyML models and IoT devices at scale, from modeling information to discovering possible combinations and benchmarking, and eventually facilitate TinyML component exchange and reuse. We present an ontology (semantic schema) for neural network models aligned with the World Wide Web Consortium (W3C) Thing Description, which semantically describes IoT devices. Furthermore, a Knowledge Graph of 23 publicly available ML models and six IoT devices were used to demonstrate our concept in three case studies, and we shared the code and examples to enhance reproducibility: https://github.com/Haoyu-R/How-to-Manage-TinyML-at-Scale

cs.AI↗

arXiv2022

Comparing Model-free and Model-based Algorithms for Offline Reinforcement Learning

Offline reinforcement learning (RL) Algorithms are often designed with environments such as MuJoCo in mind, in which the planning horizon is extremely long and no noise exists. We compare model-free, model-based, as well as hybrid offline RL approaches on various industrial benchmark (IB) datasets to test the algorithms in settings closer to real world problems, including complex noise and partially observable states. We find that on the IB, hybrid approaches face severe difficulties and that simpler algorithms, such as rollout based algorithms or model-free algorithms with simpler regularizers perform best on the datasets.

cs.LG↗

arXiv2021

Combining Sub-Symbolic and Symbolic Methods for Explainability

Similarly to other connectionist models, Graph Neural Networks (GNNs) lack transparency in their decision-making. A number of sub-symbolic approaches have been developed to provide insights into the GNN decision making process. These are first important steps on the way to explainability, but the generated explanations are often hard to understand for users that are not AI experts. To overcome this problem, we introduce a conceptual approach combining sub-symbolic and symbolic methods for human-centric explanations, that incorporate domain knowledge and causality. We furthermore introduce the notion of fidelity as a metric for evaluating how close the explanation is to the GNN's internal decision making process. The evaluation with a chemical dataset and ontology shows the explanatory value and reliability of our method.

cs.AI↗

arXiv2021

Measuring Data Quality for Dataset Selection in Offline Reinforcement Learning

Recently developed offline reinforcement learning algorithms have made it possible to learn policies directly from pre-collected datasets, giving rise to a new dilemma for practitioners: Since the performance the algorithms are able to deliver depends greatly on the dataset that is presented to them, practitioners need to pick the right dataset among the available ones. This problem has so far not been discussed in the corresponding literature. We discuss ideas how to select promising datasets and propose three very simple indicators: Estimated relative return improvement (ERI) and estimated action stochasticity (EAS), as well as a combination of the two (COI), and empirically show that despite their simplicity they can be very effectively used for dataset selection.

cs.LG↗