arXiv subjects

Artificial intelligence

Search indexed arXiv papers on artificial intelligence and machine learning, including cs.AI metadata. Follow the original manuscripts for methods, experiments and version history.

Follow this topic with RSS

At least 1,531 records · Page 85Linked to original sources

arXiv2026

BiasLab: A Multilingual Dual-Framing Framework for LLM Bias Measurement, Applied to Workplace and HR Contexts

Background: Large language models (LLMs) harbor systematic biases that are particularly consequential in workplace and HR contexts, where their outputs increasingly influence hiring, job design, and organizational decisions. Existing bias-evaluation approaches remain methodologically fragmented, limiting practitioners' ability to assess deployment risks. Objective: This study introduces BiasLab, a multilingual dual-framing framework to quantify and compare directional output-level bias in LLMs, demonstrated across six workplace and HR-relevant topics. Methods: BiasLab combines mirrored affirmative and reverse prompt pairs, randomized wrapper perturbations, fixed-choice response constraints, and polarity-aligned scoring. Ten LLMs were evaluated across six topics (gender in leadership, employment gap candidates, age in hiring, remote versus office work, four-day versus five-day work weeks, and AI-assisted versus human-only hiring), spanning 12 languages and 30 iterations per framing direction, yielding 43,200 responses. Results: All ten models showed consistent directional preferences across every topic. A recurring asymmetric pattern emerged in which models rejected disfavored claims more strongly than they endorsed their opposites, a distinction invisible to single-frame designs. Conclusions: BiasLab provides a standardized, reproducible instrument for measuring directional preferences across models. Whether a preference constitutes bias in a fairness sense is topic-dependent: for protected attributes such as gender and age it maps onto equal-employment standards, whereas elsewhere it is better described as systematic preference. The framework lets organizations compare and vet models before adopting them for hiring.

cs.CL↗

arXiv2026

On Sample-Efficient Generalized Planning via Learned Transition Models

Generalized planning studies the construction of solution strategies that generalize across families of planning problems sharing a common domain model, formally defined by a transition function $γ: S \times A \rightarrow S$. Classical approaches achieve such generalization through symbolic abstractions and explicit reasoning over $γ$. In contrast, recent Transformer-based planners, such as PlanGPT and Plansformer, largely cast generalized planning as direct action-sequence prediction, bypassing explicit transition modeling. While effective on in-distribution instances, these approaches typically require large datasets and model sizes, and often suffer from state drift in long-horizon settings due to the absence of explicit world-state evolution. In this work, we formulate generalized planning as a transition-model learning problem, in which a neural model explicitly approximates the successor-state function $\hatγ \approx γ$ and generates plans by rolling out symbolic state trajectories. Instead of predicting actions directly, the model autoregressively predicts intermediate world states, thereby learning the domain dynamics as an implicit world model. To study size-invariant generalization and sample efficiency, we systematically evaluate multiple state representations and neural architectures, including relational graph encodings. Our results show that learning explicit transition models yields higher out-of-distribution satisficing-plan success than direct action-sequence prediction in multiple domains, while achieving these gains with significantly fewer training instances and smaller models. This is an extended version of a short paper accepted at ICAPS 2026 under the same title.

cs.AI↗

arXiv2025

Doing More with Less: A Survey on Routing Strategies for Resource Optimisation in Large Language Model-Based Systems

Large Language Model (LLM)-based systems, i.e. interconnected elements that include an LLM as a central component, such as conversational agents, are usually designed with monolithic, static architectures that rely on a single, general-purpose LLM to handle all user queries. However, these systems may be inefficient as different queries may require different levels of reasoning, domain knowledge or pre-processing. While generalist LLMs (e.g. GPT-4o, Claude-Sonnet) perform well across a wide range of tasks, they may incur significant financial, energy and computational costs. These costs may be disproportionate for simpler queries, resulting in unnecessary resource utilisation. A routing mechanism can therefore be employed to route queries to more appropriate components, such as smaller or specialised models, thereby improving efficiency and optimising resource consumption. This survey aims to provide a comprehensive overview of routing strategies in LLM-based systems. Specifically, it reviews when, why, and how routing should be integrated into LLM pipelines to improve efficiency, scalability, and performance. We define the objectives to optimise, such as cost minimisation and performance maximisation, and discuss the timing of routing within the LLM workflow, whether it occurs before or after generation. We also detail the various implementation strategies, including similarity-based, supervised, reinforcement learning-based, and generative methods. Practical considerations such as industrial applications and current limitations are also examined, like standardising routing experiments, accounting for non-financial costs, and designing adaptive strategies. By formalising routing as a performance-cost optimisation problem, this survey provides tools and directions to guide future research and development of adaptive low-cost LLM-based systems.

cs.AI↗

arXiv2026

How to DP-fy Your Data: A Practical Guide to Generating Synthetic Data With Differential Privacy

High quality data is needed to unlock the full potential of AI for end users. However finding new sources of such data is getting harder: most publicly-available human generated data will soon have been used. Additionally, publicly available data often is not representative of users of a particular system -- for example, a research speech dataset of contractors interacting with an AI assistant will likely be more homogeneous, well articulated and self-censored than real world commands that end users will issue. Therefore unlocking high-quality data grounded in real user interactions is of vital interest. However, the direct use of user data comes with significant privacy risks. Differential Privacy (DP) is a well established framework for reasoning about and limiting information leakage, and is a gold standard for protecting user privacy. The focus of this work, \emph{Differentially Private Synthetic data}, refers to synthetic data that preserves the overall trends of source data,, while providing strong privacy guarantees to individuals that contributed to the source dataset. DP synthetic data can unlock the value of datasets that have previously been inaccessible due to privacy concerns and can replace the use of sensitive datasets that previously have only had rudimentary protections like ad-hoc rule-based anonymization. In this paper we explore the full suite of techniques surrounding DP synthetic data, the types of privacy protections they offer and the state-of-the-art for various modalities (image, tabular, text and decentralized). We outline all the components needed in a system that generates DP synthetic data, from sensitive data handling and preparation, to tracking the use and empirical privacy testing. We hope that work will result in increased adoption of DP synthetic data, spur additional research and increase trust in DP synthetic data approaches.

cs.CR↗

arXiv2025

Interactive-KBQA: Multi-Turn Interactions for Knowledge Base Question Answering with Large Language Models

This study explores the realm of knowledge base question answering (KBQA). KBQA is considered a challenging task, particularly in parsing intricate questions into executable logical forms. Traditional semantic parsing (SP)-based methods require extensive data annotations, which result in significant costs. Recently, the advent of few-shot in-context learning, powered by large language models (LLMs), has showcased promising capabilities. However, fully leveraging LLMs to parse questions into logical forms in low-resource scenarios poses a substantial challenge. To tackle these hurdles, we introduce Interactive-KBQA, a framework designed to generate logical forms through direct interaction with knowledge bases (KBs). Within this framework, we have developed three generic APIs for KB interaction. For each category of complex question, we devised exemplars to guide LLMs through the reasoning processes. Our method achieves competitive results on the WebQuestionsSP, ComplexWebQuestions, KQA Pro, and MetaQA datasets with a minimal number of examples (shots). Importantly, our approach supports manual intervention, allowing for the iterative refinement of LLM outputs. By annotating a dataset with step-wise reasoning processes, we showcase our model's adaptability and highlight its potential for contributing significant enhancements to the field.

cs.CL↗

arXiv2026

PromptRad: Knowledge-Enhanced Multi-Label Prompt-Tuning for Low-Resource Radiology Report Labeling

Automatic report labeling facilitates the identification of clinical findings from unstructured text and enables large-scale annotation for medical imaging research. Existing rule-based labelers struggle with the diverse descriptions in clinical reports, while fine-tuning pre-trained language models (PLMs) requires large amounts of labeled data that are often unavailable in clinical settings. In this paper, we propose PromptRad, a knowledge-enhanced multi-label \textbf{prompt}-tuning approach for \textbf{rad}iology report labeling under low-resource settings. PromptRad reformulates multi-label classification as masked language modeling and incorporates synonyms from the UMLS Metathesaurus into a multi-word verbalizer to enrich category representations. By fine-tuning the PLM without additional classification layers, PromptRad requires substantially less labeled data than conventional fine-tuning. Experiments on liver CT (computed tomography) reports show that PromptRad outperforms dictionary-based and fine-tuning baselines with only 32 labeled training examples, and achieves competitive performance with GPT-4 despite using a much smaller model. Further analysis demonstrates that PromptRad captures complex negation patterns more effectively than existing methods, making it a promising solution for report labeling in data-scarce clinical scenarios. Our code is available at https://github.com/ila-lab/PromptRad.

cs.CL↗

arXiv2026

Valid $\ne$ Necessary: Diagnosing Latent Inefficiency in Chain-of-Thought

Chain-of-Thought (CoT) prompting has significantly advanced the reasoning capabilities of Large Language Models (LLMs), yet it often incurs substantial computational costs due to over-reasoning: the generation of redundant, verbose, or irrelevant steps. While existing reasoning step evaluators effectively detect logical fallacies and factual errors, our analysis reveals a critical blind spot: they fail to penalize valid but inefficient reasoning steps that inflate token usage without contributing to the solution. To systematically diagnose this limitation, we introduce RIV-GSM8K, a diagnostic benchmark injected with five distinct types of inefficiencies, including circular reasoning and excessive decomposition. Diagnostic experiments reveal that state-of-the-art evaluators struggle to distinguish these inefficiencies from necessary reasoning. To address this gap, we propose CAID (Context-Aware Information Density), a training-free metric grounded in information theory that identifies low-utility steps. To validate the metric's practical utility, we apply it within PACE, a post-hoc compression strategy. Additional control experiments show that the gains of PACE are not explained by trivial pruning: compared with random step removal and PRM-based compression baselines, it preserves accuracy at substantially higher compression rates. Empirical results on GSM8K, StrategyQA, and ARC-Challenge demonstrate that PACE reduces token consumption by 31-53% while maintaining accuracy, confirming that CAID successfully distills informational froth from reasoning chains without compromising deductive validity.

cs.AI↗

arXiv2026

Lightning Fast Matching Dependency Discovery with Desbordante

Matching dependency is a generalization of the functional dependency concept, which allows users to apply custom similarity functions for matching individual attributes. Matching dependencies have a wide range of applications for solving various data quality problems, such as entity resolution, data deduplication, data integration, schema matching, and many more. However, their discovery is a very computationally intensive problem, which limits their practical application. In this paper, we describe a number of optimization techniques for HyMD - currently the state-of-the-art algorithm for the discovery of matching dependencies. These optimizations belong to both technical and scientific domains. The most important of them are: 1) a new sampling technique, 2) a faster generalization lookup technique, and 3) an improved representation of a dependency. The first one aims to raise the efficiency of inference from record pairs, while the last two are designed to speed up lattice-related operations. To evaluate our optimizations, we implemented our version of HyMD in Desbordante, an open-source high-performance data profiler. Experiments demonstrated that they allow for a speedup of more than 40x over the state-of-the-art implementation on average, reaching a speedup greater than 170x in some cases. Finally, the improved version of HyMD is ready to use by anyone. It comes with bidirectional Python integration, which allows calling the C++ algorithm implementation from Python programs while allowing users to supply their custom matching functions.

cs.DB↗

arXiv2025

Dubito Ergo Sum: Exploring AI Ethics

We paraphrase Descartes' famous dictum in the area of AI ethics where the "I doubt and therefore I am" is suggested as a necessary aspect of morality. Therefore AI, which cannot doubt itself, cannot possess moral agency. Of course, this is not the end of the story. We explore various aspects of the human mind that substantially differ from AI, which includes the sensory grounding of our knowing, the act of understanding, and the significance of being able to doubt ourselves. The foundation of our argument is the discipline of ethics, one of the oldest and largest knowledge projects of human history, yet, we seem only to be beginning to get a grasp of it. After a couple of thousand years of studying the ethics of humans, we (humans) arrived at a point where moral psychology suggests that our moral decisions are intuitive, and all the models from ethics become relevant only when we explain ourselves. This recognition has a major impact on what and how we can do regarding AI ethics. We do not offer a solution, we explore some ideas and leave the problem open, but we hope somewhat better understood than before our study.

cs.AI↗

arXiv2025

A Simplified Retriever to Improve Accuracy of Phenotype Normalizations by Large Language Models

Large language models (LLMs) have shown improved accuracy in phenotype term normalization tasks when augmented with retrievers that suggest candidate normalizations based on term definitions. In this work, we introduce a simplified retriever that enhances LLM accuracy by searching the Human Phenotype Ontology (HPO) for candidate matches using contextual word embeddings from BioBERT without the need for explicit term definitions. Testing this method on terms derived from the clinical synopses of Online Mendelian Inheritance in Man (OMIM), we demonstrate that the normalization accuracy of a state-of-the-art LLM increases from a baseline of 62.3% without augmentation to 90.3% with retriever augmentation. This approach is potentially generalizable to other biomedical term normalization tasks and offers an efficient alternative to more complex retrieval methods.

cs.CL↗

arXiv2025

Addressing Moral Uncertainty using Large Language Models for Ethical Decision-Making

We present an ethical decision-making framework that refines a pre-trained reinforcement learning (RL) model using a task-agnostic ethical layer. Following initial training, the RL model undergoes ethical fine-tuning, where human feedback is replaced by feedback generated from a large language model (LLM). The LLM embodies consequentialist, deontological, virtue, social justice, and care ethics as moral principles to assign belief values to recommended actions during ethical decision-making. An ethical layer aggregates belief scores from multiple LLM-derived moral perspectives using Belief Jensen-Shannon Divergence and Dempster-Shafer Theory into probability scores that also serve as the shaping reward, steering the agent toward choices that align with a balanced ethical framework. This integrated learning framework helps the RL agent navigate moral uncertainty in complex environments and enables it to make morally sound decisions across diverse tasks. Our approach, tested across different LLM variants and compared with other belief aggregation techniques, demonstrates improved consistency, adaptability, and reduced reliance on handcrafted ethical rewards. This method is especially effective in dynamic scenarios where ethical challenges arise unexpectedly, making it well-suited for real-world applications.

cs.CY↗

arXiv2026

Execution-bound advisory automation for agentic AI: a reproducible AIBOM-driven CSAF-VEX framework

A protocol driven framework is presented that binds SBOM and AIBOM artefacts to deterministic environment capture and structured runtime telemetry. Exploitability is computed from declared artefacts, observed activation conditions, and enforced execution policies. CSAF VEX advisories are generated from combined static and runtime evidence, cryptographically signed, and validated through deterministic replay. Evaluation uses approximately 10000 component entries across synthetic Agentic AI workloads 50 to 5000 components, incorporating OSV, GitHub Advisory, KEV, and EPSS datasets.

cs.SE↗

arXiv2026

From Black Box to Glass Box: Cross-Model ASR Disagreement to Prioto Review in Ambient AI Scribe Documentation

Ambient AI "scribe" systems promise to reduce clinical documentation burden, but automatic speech recognition (ASR) errors can remain unnoticed without careful review, and high-quality human reference transcripts are often unavailable for calibrating uncertainty. We investigate whether cross-model disagreement among heterogeneous ASR systems can act as a reference-free uncertainty signal to prioritize human verification in medical transcription workflows. Using 50 publicly available medical education audio clips (8 h 14 min), we transcribed each clip with eight ASR systems spanning commercial APIs and open-source engines. We aligned multi-model outputs, built consensus pseudo-references, and quantified token-level agreement using a majority-strength metric; we further characterized disagreements by type (content vs. punctuation/formatting) and assessed per-model agreement via leave-one-model-out (jackknife) consensus scoring. Inter-model reliability was low (ICC[2,1] = 0.131), indicating heterogeneous failure modes across systems. Across 76,398 evaluated token positions, 72.1% showed near-unanimous agreement (7-8 models), while 2.5% fell into high-risk bands (0-3 models), with high-risk mass varying from 0.7% to 11.4% across accent groups. Low-agreement regions were enriched for content disagreements, with the content fraction increasing from 53.9% to 73.9% across quintiles of high-risk mass. These results suggest that cross-model disagreement provides a sparse, localizable signal that can surface potentially unreliable transcript spans without human-verified references, enabling targeted review; clinical accuracy of flagged regions remains to be established.

cs.SD↗

arXiv2020

Adaptive Prior Selection for Repertoire-based Online Adaptation in Robotics

Repertoire-based learning is a data-efficient adaptation approach based on a two-step process in which (1) a large and diverse set of policies is learned in simulation, and (2) a planning or learning algorithm chooses the most appropriate policies according to the current situation (e.g., a damaged robot, a new object, etc.). In this paper, we relax the assumption of previous works that a single repertoire is enough for adaptation. Instead, we generate repertoires for many different situations (e.g., with a missing leg, on different floors, etc.) and let our algorithm selects the most useful prior. Our main contribution is an algorithm, APROL (Adaptive Prior selection for Repertoire-based Online Learning) to plan the next action by incorporating these priors when the robot has no information about the current situation. We evaluate APROL on two simulated tasks: (1) pushing unknown objects of various shapes and sizes with a robotic arm and (2) a goal reaching task with a damaged hexapod robot. We compare with "Reset-free Trial and Error" (RTE) and various single repertoire-based baselines. The results show that APROL solves both the tasks in less interaction time than the baselines. Additionally, we demonstrate APROL on a real, damaged hexapod that quickly learns to pick compensatory policies to reach a goal by avoiding obstacles in the path.

cs.RO↗

arXiv2022

Affect-Driven Modelling of Robot Personality for Collaborative Human-Robot Interactions

Collaborative interactions require social robots to adapt to the dynamics of human affective behaviour. Yet, current approaches for affective behaviour generation in robots focus on instantaneous perception to generate a one-to-one mapping between observed human expressions and static robot actions. In this paper, we propose a novel framework for personality-driven behaviour generation in social robots. The framework consists of (i) a hybrid neural model for evaluating facial expressions and speech, forming intrinsic affective representations in the robot, (ii) an Affective Core, that employs self-organising neural models to embed robot personality traits like patience and emotional actuation, and (iii) a Reinforcement Learning model that uses the robot's affective appraisal to learn interaction behaviour. For evaluation, we conduct a user study (n = 31) where the NICO robot acts as a proposer in the Ultimatum Game. The effect of robot personality on its negotiation strategy is witnessed by participants, who rank a patient robot with high emotional actuation higher on persistence, while an inert and impatient robot higher on its generosity and altruistic behaviour.

cs.RO↗

arXiv2024

Hierarchical Path-planning from Speech Instructions with Spatial Concept-based Topometric Semantic Mapping

Assisting individuals in their daily activities through autonomous mobile robots, especially for users without specialized knowledge, is crucial. Specifically, the capability of robots to navigate to destinations based on human speech instructions is essential. While robots can take different paths to the same goal, the shortest path is not always the best. A preferred approach is to accommodate waypoint specifications flexibly, planning an improved alternative path, even with detours. Additionally, robots require real-time inference capabilities. This study aimed to realize a hierarchical spatial representation using a topometric semantic map and path planning with speech instructions, including waypoints. This paper presents Spatial Concept-based Topometric Semantic Mapping for Hierarchical Path Planning (SpCoTMHP), integrating place connectivity. This approach offers a novel integrated probabilistic generative model and fast approximate inference across hierarchy levels. A formulation based on control as probabilistic inference theoretically supports the proposed path planning algorithm. We conducted experiments in home environments using the Toyota Human Support Robot on the SIGVerse simulator and in a lab-office environment with the real robot, Albert. Users issued speech commands specifying the waypoint and goal, such as "Go to the bedroom via the corridor." Navigation experiments using speech instructions with a waypoint demonstrated a performance improvement of SpCoTMHP over the baseline hierarchical path planning method with heuristic path costs (HPP-I), in terms of the weighted success rate at which the robot reaches the closest target and passes the correct waypoints, by 0.590. The computation time was significantly accelerated by 7.14 seconds with SpCoTMHP compared to baseline HPP-I in advanced tasks.

cs.RO↗

arXiv2024

Exploring Large Language Models to Facilitate Variable Autonomy for Human-Robot Teaming

In a rapidly evolving digital landscape autonomous tools and robots are becoming commonplace. Recognizing the significance of this development, this paper explores the integration of Large Language Models (LLMs) like Generative pre-trained transformer (GPT) into human-robot teaming environments to facilitate variable autonomy through the means of verbal human-robot communication. In this paper, we introduce a novel framework for such a GPT-powered multi-robot testbed environment, based on a Unity Virtual Reality (VR) setting. This system allows users to interact with robot agents through natural language, each powered by individual GPT cores. By means of OpenAI's function calling, we bridge the gap between unstructured natural language input and structure robot actions. A user study with 12 participants explores the effectiveness of GPT-4 and, more importantly, user strategies when being given the opportunity to converse in natural language within a multi-robot environment. Our findings suggest that users may have preconceived expectations on how to converse with robots and seldom try to explore the actual language and cognitive capabilities of their robot collaborators. Still, those users who did explore where able to benefit from a much more natural flow of communication and human-like back-and-forth. We provide a set of lessons learned for future research and technical implementations of similar systems.

cs.HC↗

arXiv2024

Incremental Learning of Humanoid Robot Behavior from Natural Interaction and Large Language Models

Natural-language dialog is key for intuitive human-robot interaction. It can be used not only to express humans' intents, but also to communicate instructions for improvement if a robot does not understand a command correctly. Of great importance is to endow robots with the ability to learn from such interaction experience in an incremental way to allow them to improve their behaviors or avoid mistakes in the future. In this paper, we propose a system to achieve incremental learning of complex behavior from natural interaction, and demonstrate its implementation on a humanoid robot. Building on recent advances, we present a system that deploys Large Language Models (LLMs) for high-level orchestration of the robot's behavior, based on the idea of enabling the LLM to generate Python statements in an interactive console to invoke both robot perception and action. The interaction loop is closed by feeding back human instructions, environment observations, and execution results to the LLM, thus informing the generation of the next statement. Specifically, we introduce incremental prompt learning, which enables the system to interactively learn from its mistakes. For that purpose, the LLM can call another LLM responsible for code-level improvements of the current interaction based on human feedback. The improved interaction is then saved in the robot's memory, and thus retrieved on similar requests. We integrate the system in the robot cognitive architecture of the humanoid robot ARMAR-6 and evaluate our methods both quantitatively (in simulation) and qualitatively (in simulation and real-world) by demonstrating generalized incrementally-learned knowledge.

cs.RO↗