arXiv subjects

William Yeoh

Publications and source records attributed to William Yeoh.

At least 19 recordsLinked to original sources

arXiv2026

ROPE: Routed Origin Policy Enforcement against Indirect Prompt Injection

Indirect prompt injection (IPI) plants instructions in the content a tool-using LLM agent reads, steering the agent into harmful tool calls. The strongest defenses are system-level, leveraging techniques such as task-conditional tool screening to prevent execution of malicious tools, and information-flow control to avoid tool execution with untrusted parameters. However, as agents grow more capable, users delegate more to automation. Consequently, tool execution sequences and parameter values are increasingly determined at runtime and cannot be reliably screened from solely user's query without significant utility loss. We present ROPE (Routed Origin Policy Enforcement), which is anchored in a structural notion of trust: a value may reach a state-changing tool only if it traces unforgeably to the user, a source the user explicitly named, or the user's own authoritative records. Enforcement is then a deterministic origin check over an audited set of sensitive tool parameters, and the only reliance on a language model involves solely the trusted user request, out of the attacker's reach. Our approach admits two provable guarantees: 1) at every step of a trajectory, no value whose only origin is attacker-writable content reaches an origin-guarded parameter, and 2) no rewording of an injection changes an admission decision. We evaluate across four agent models on open-ended agent suites, ROPE holds attack success rate to 1.6--2.6\% while retaining 82--100\% of undefended clean utility, significantly exceeding state-of-the-art system-level defenses in utility while attaining comparable or better security. Further, we show that optimizing the injection against ROPE is largely ineffective, while long-horizon attacks that defeat prior system-level defenses achieve zero success rate. Our code and logs are available at https://github.com/xhOwenMa/ROPE .

cs.CR↗

arXiv2026

From Efficiency to Leakage -- Privacy Backdoor in Federated Language Model Fine-Tuning

Federated learning (FL) enables multiple parties to collaboratively fine-tune language models for domain-specific tasks without sharing raw data. Since full model fine-tuning is often prohibitively expensive for FL clients, parameter-efficient fine-tuning (PEFT) has become the de facto approach in practice, freezing the base model and training only a small set of adapters. In this paper, we show that a malicious parameter server can stealthily corrupt a PEFT adapter into a privacy backdoor that implicitly memorizes the client's training samples as isolated per-sample parameter updates stored in separate neurons, without degrading model utility. Concretely, our attack, NeuroImprint, assigns a dedicated memorization neuron to each training sample and constrains that each neuron is updated at most once along the local fine-tuning trajectory. This design mitigates both cross-sample collisions and cross-step mixing introduced by large local batches and stateful optimizers (e.g., Adam/AdamW) in language-model fine-tuning. After fine-tuning, the resulting isolated per-sample updates can be analytically inverted in closed form to recover text embeddings, which are then deterministically mapped back to token sequences. To understand the generality of our method, we implemented NeuroImprint on multiple language models (BERT, GPT-2, Qwen2, and Llama3.2) and evaluated it across four fine-tuning datasets spanning diverse domains. The results demonstrate that our attack can reconstruct 59% to 79% of all finetuning samples with high semantic fidelity.

cs.CR↗

arXiv2026

LifeSentence: Language models can encode human life course trajectories from longitudinal panel data

Forecasting human life outcomes is important to gain insights into how individuals attain long and healthy lives. Conventional statistical approaches yield limited accuracy, potentially due to discarding the sequential structure of the life course. Modern methods such as transformer architectures require large scale training data that most longitudinal panel studies lack. Here we introduce LifeSentence, a model for life-course reasoning that bridges large language models with longitudinal panel data. By representing each life event as a structured natural-language record and instruction-tuning a pretrained 24-billion-parameter language model across an 18-task evaluation taxonomy spanning prediction, robustness and reasoning, LifeSentence supplements panel data with distributional knowledge already encoded during pretraining. Trained on approximately 65,000 individuals from the German Socio-Economic Panel - roughly 45 times fewer than prior transformer-based approaches - LifeSentence outperforms classical and deep learning baselines across all task families, achieving a threefold improvement in joint event-and-timing prediction from best baselines and 91.2% Kendall's tau when reconstructing chronological order from timestamp-stripped event sets. Without explicit supervision, the model recovers documented patterns of social stratification, including the education premium, the gender wage gap and the motherhood penalty, from discrete event sequences alone. A natural-language interface further enables qualitatively new research queries, such as connecting an early-life history to a specified late-life endpoint, establishing LifeSentence as both a predictive tool and a probe for counterfactual exploration of human biographies.

cs.CL↗

arXiv2026

COSMOS: Model-Agnostic Personalized Federated Learning with Clustered Server Models and Pseudo-Label-Only Communication

Federated learning (FL) in heterogeneous environments remains challenging because client models often differ in both architecture and data distribution. While recent approaches attempt to address this challenge through client clustering and knowledge distillation, simultaneously handling architectural and statistical heterogeneity remains difficult. We introduce COSMOS, a model-agnostic framework that enables server-side personalization using only pseudo-label communication. Clients train local models and predict on the public data; the server clusters clients by prediction similarity, trains a cluster-specific model for each group using its own compute, and distills the resulting models back to clients. We provide the first theoretical analysis showing that distillation from the learned cluster models can yield exponential personalization risk contraction, going beyond the convergence-to-stationarity guarantees typically provided in model-agnostic FL. Experiments across benchmarks demonstrate that COSMOS consistently outperforms all model-agnostic FL baselines while remaining competitive with state-of-the-art personalized FL methods. More broadly, our results highlight personalized server-side learning with pseudo-labels as a promising paradigm for scalable and model-agnostic federated learning in highly heterogeneous environments.

cs.LG↗

arXiv2026

Protecting Language Models Against Unauthorized Distillation through Trace Rewriting

Knowledge distillation is a widely adopted technique for transferring capabilities from LLMs to smaller, more efficient student models. However, unauthorized use of knowledge distillation takes unfair advantage of the considerable effort and cost put into developing frontier models. We investigate methods for modifying teacher-generated reasoning traces to achieve two objectives that deter unauthorized distillation: (1) \emph{anti-distillation}, or degrading the training usefulness of query responses, and (2) \emph{API watermarking}, which embeds verifiable signatures in student models. We introduce several approaches for dynamically rewriting a teacher's reasoning outputs while preserving answer correctness and semantic coherence. Two of these leverage the rewriting capabilities of LLMs, while others use gradient-based techniques. Our experiments show that a simple instruction-based rewriting approach achieves a strong anti-distillation effect while maintaining or even improving teacher performance. Furthermore, we show that our rewriting approach also enables embedding watermarks that can be reliably detected with essentially no false alarms. Our code is available at https://github.com/xhOwenMa/trace-rewriting.

cs.AI↗

arXiv2026

Argumentative Human-AI Decision-Making: Toward AI Agents That Reason With Us, Not For Us

Computational argumentation offers formal frameworks for transparent, verifiable reasoning but has traditionally been limited by its reliance on domain-specific information and extensive feature engineering. In contrast, LLMs excel at processing unstructured text, yet their opaque nature makes their reasoning difficult to evaluate and trust. We argue that the convergence of these fields will lay the foundation for a new paradigm: Argumentative Human-AI Decision-Making. We analyze how the synergy of argumentation framework mining, argumentation framework synthesis, and argumentative reasoning enables agents that do not just justify decisions, but engage in dialectical processes where decisions are contestable and revisable -- reasoning with humans rather than for them. This convergence of computational argumentation and LLMs is essential for human-aware, trustworthy AI in high-stakes domains.

cs.AI↗

arXiv2026

UniRel: Relation-Centric Knowledge Graph Question Answering with RL-Tuned LLM Reasoning

Knowledge Graph Question Answering (KGQA) has largely focused on entity-centric queries that return a single answer entity. However, many real-world questions are inherently relational, aiming to understand how entities are associated rather than which entity satisfies a query. In this work, we introduce relation-centric KGQA, a complementary setting in which the answer is a subgraph that represents the semantic relations among entities. The main challenge lies in the abundance of candidate subgraphs, where trivial or overly common connections often obscure the identification of unique and informative answers. To tackle this, we propose UniRel, a unified modular framework that combines a subgraph retriever with an LLM fine-tuned using reinforcement learning. The framework uses a reward function to prefer compact and specific subgraphs with informative relations and low-degree intermediate entities. Experiments show that UniRel improves connectivity and reward over Prompting baselines and generalizes well to unseen entities and relations. Moreover, UniRel can be applied to conventional entity-centric KGQA, achieving competitive or improved performance in several settings.

cs.AI↗

arXiv2026

Past-Discounting is Key for Learning Markovian Fairness with Long Horizons

Fairness is an important consideration for dynamic resource allocation in multi-agent systems. Many existing methods treat fairness as a one-shot problem without considering temporal dynamics, which misses the nuances of accumulating inequalities over time. Recent approaches overcome this limitation by tracking allocations over time, assuming perfect recall of all past utilities. While the former neglects long-term equity, the latter introduces a critical challenge: the augmented state space required to track cumulative utilities grows unboundedly with time, hindering the scalability and convergence of learning algorithms. Motivated by behavioral insights that human fairness judgments discount distant events, we introduce a framework for temporal fairness that incorporates past-discounting into the learning problem. This approach offers a principled interpolation between instantaneous and perfect-recall fairness. Our central contribution is a past-discounted framework for memory tracking and a theoretical analysis of fairness memories, showing past-discounting guarantees a bounded, horizon-independent state space, a property that we prove perfect-recall methods lack. This result unlocks the ability to learn fair policies tractably over arbitrarily long horizons. We formalize this framework, demonstrate its necessity with experiments showing that perfect recall fails where past-discounting succeeds, and provide a clear path toward building scalable and equitable resource allocation systems.

cs.AI↗

arXiv2025

EcoLoRA: Communication-Efficient Federated Fine-Tuning of Large Language Models

To address data locality and privacy restrictions, Federated Learning (FL) has recently been adopted to fine-tune large language models (LLMs), enabling improved performance on various downstream tasks without requiring aggregated data. However, the repeated exchange of model updates in FL can result in prohibitively high communication costs, hindering the distributed learning process. To address this challenge, we propose EcoLoRA, a novel communication-efficient federated fine-tuning framework for LLMs. Leveraging the modular structure, we propose a round-robin segment sharing scheme, where each client uploads only a complementary LoRA segment per round to reduce network bandwidth. It is further combined with adaptive sparsification methods tailored to LoRA's training dynamics and lossless encoding techniques. We conduct extensive evaluations on both question-answering and value-alignment tasks across multiple datasets and models. The results show that EcoLoRA significantly reduces communication overhead without compromising performance. For instance, it reduces communication time by up to 79% and total training time by up to 65%.

cs.DC↗

arXiv2025

A General Incentives-Based Framework for Fairness in Multi-agent Resource Allocation

We introduce the General Incentives-based Framework for Fairness (GIFF), a novel approach for fair multi-agent resource allocation that infers fair decision-making from standard value functions. In resource-constrained settings, agents optimizing for efficiency often create inequitable outcomes. Our approach leverages the action-value (Q-)function to balance efficiency and fairness without requiring additional training. Specifically, our method computes a local fairness gain for each action and introduces a counterfactual advantage correction term to discourage over-allocation to already well-off agents. This approach is formalized within a centralized control setting, where an arbitrator uses the GIFF-modified Q-values to solve an allocation problem. Empirical evaluations across diverse domains, including dynamic ridesharing, homelessness prevention, and a complex job allocation task-demonstrate that our framework consistently outperforms strong baselines and can discover far-sighted, equitable policies. The framework's effectiveness is supported by a theoretical foundation; we prove its fairness surrogate is a principled lower bound on the true fairness improvement and that its trade-off parameter offers monotonic tuning. Our findings establish GIFF as a robust and principled framework for leveraging standard reinforcement learning components to achieve more equitable outcomes in complex multi-agent systems.

cs.MA↗

arXiv2025

Mind the Gaps: Auditing and Reducing Group Inequity in Large-Scale Mobility Prediction

Next location prediction underpins a growing number of mobility, retail, and public-health applications, yet its societal impacts remain largely unexplored. In this paper, we audit state-of-the-art mobility prediction models trained on a large-scale dataset, highlighting hidden disparities based on user demographics. Drawing from aggregate census data, we compute the difference in predictive performance on racial and ethnic user groups and show a systematic disparity resulting from the underlying dataset, resulting in large differences in accuracy based on location and user groups. To address this, we propose Fairness-Guided Incremental Sampling (FGIS), a group-aware sampling strategy designed for incremental data collection settings. Because individual-level demographic labels are unavailable, we introduce Size-Aware K-Means (SAKM), a clustering method that partitions users in latent mobility space while enforcing census-derived group proportions. This yields proxy racial labels for the four largest groups in the state: Asian, Black, Hispanic, and White. Built on these labels, our sampling algorithm prioritizes users based on expected performance gains and current group representation. This method incrementally constructs training datasets that reduce demographic performance gaps while preserving overall accuracy. Our method reduces total disparity between groups by up to 40\% with minimal accuracy trade-offs, as evaluated on a state-of-art MetaPath2Vec model and a transformer-encoder model. Improvements are most significant in early sampling stages, highlighting the potential for fairness-aware strategies to deliver meaningful gains even in low-resource settings. Our findings expose structural inequities in mobility prediction pipelines and demonstrate how lightweight, data-centric interventions can improve fairness with little added complexity, especially for low-data applications.

cs.LG↗

arXiv2025

Goal Recognition Design for General Behavioral Agents using Machine Learning

Goal recognition design (GRD) aims to make limited modifications to decision-making environments to make it easier to infer the goals of agents acting within those environments. Although various research efforts have been made in goal recognition design, existing approaches are computationally demanding and often assume that agents are (near-)optimal in their decision-making. To address these limitations, we leverage machine learning methods for goal recognition design that can both improve run-time efficiency and account for agents with general behavioral models. Following existing literature, we use worst-case distinctiveness (wcd) as a measure of the difficulty in inferring the goal of an agent in a decision-making environment. Our approach begins by training a machine learning model to predict the wcd for a given environment and the agent behavior model. We then propose a gradient-based optimization framework that accommodates various constraints to optimize decision-making environments for enhanced goal recognition. Through extensive simulations, we demonstrate that our approach outperforms existing methods in reducing wcd and enhances runtime efficiency. Moreover, our approach also adapts to settings in which existing approaches do not apply, such as those involving flexible budget constraints, more complex environments, and suboptimal agent behavior. Finally, we conducted human-subject experiments that demonstrate that our method creates environments that facilitate efficient goal recognition from human decision-makers.

cs.AI↗

arXiv2025

On Generating Monolithic and Model Reconciling Explanations in Probabilistic Scenarios

Explanation generation frameworks aim to make AI systems' decisions transparent and understandable to human users. However, generating explanations in uncertain environments characterized by incomplete information and probabilistic models remains a significant challenge. In this paper, we propose a novel framework for generating probabilistic monolithic explanations and model reconciling explanations. Monolithic explanations provide self-contained reasons for an explanandum without considering the agent receiving the explanation, while model reconciling explanations account for the knowledge of the agent receiving the explanation. For monolithic explanations, our approach integrates uncertainty by utilizing probabilistic logic to increase the probability of the explanandum. For model reconciling explanations, we propose a framework that extends the logic-based variant of the model reconciliation problem to account for probabilistic human models, where the goal is to find explanations that increase the probability of the explanandum while minimizing conflicts between the explanation and the probabilistic human model. We introduce explanatory gain and explanatory power as quantitative metrics to assess the quality of these explanations. Further, we present algorithms that exploit the duality between minimal correction sets and minimal unsatisfiable sets to efficiently compute both types of explanations in probabilistic contexts. Extensive experimental evaluations on various benchmarks demonstrate the effectiveness and scalability of our approach in generating explanations under uncertainty.

cs.AI↗

arXiv2025

TRACE-CS: A Hybrid Logic-LLM System for Explainable Course Scheduling

We present TRACE-CS, a novel hybrid system that combines symbolic reasoning with large language models (LLMs)to address contrastive queries in course scheduling problems. TRACE-CS leverages logic-based techniques to encode scheduling constraints and generate provably correct explanations, while utilizing an LLM to process natural language queries and refine logical explanations into user friendly responses. This system showcases how combining symbolic KR methods with LLMs creates explainable AI agents that balance logical correctness with natural language accessibility, addressing a fundamental challenge in deployed scheduling systems.

cs.AI↗

arXiv2025

FairVizARD: A Visualization System for Assessing Multi-Party Fairness of Ride-Sharing Matching Algorithms

There is growing interest in algorithms that match passengers with drivers in ride-sharing problems and their fairness for the different parties involved (passengers, drivers, and ride-sharing companies). Researchers have proposed various fairness metrics for matching algorithms, but it is often unclear how one should balance the various parties' fairness, given that they are often in conflict. We present FairVizARD, a visualization-based system that aids users in evaluating the fairness of ride-sharing matching algorithms. FairVizARD presents the algorithms' results by visualizing relevant spatio-temporal information using animation and aggregated information in charts. FairVizARD also employs efficient techniques for visualizing a large amount of information in a user friendly manner, which makes it suitable for real-world settings. We conduct our experiments on a real-world large-scale taxi dataset and, through user studies and an expert interview, we show how users can use FairVizARD not only to evaluate the fairness of matching algorithms but also to expand on their notions of fairness.

cs.HC↗

arXiv2025

A Methodology for Incompleteness-Tolerant and Modular Gradual Semantics for Argumentative Statement Graphs

Gradual semantics (GS) have demonstrated great potential in argumentation, in particular for deploying quantitative bipolar argumentation frameworks (QBAFs) in a number of real-world settings, from judgmental forecasting to explainable AI. In this paper, we provide a novel methodology for obtaining GS for statement graphs, a form of structured argumentation framework, where arguments and relations between them are built from logical statements. Our methodology differs from existing approaches in the literature in two main ways. First, it naturally accommodates incomplete information, so that arguments with partially specified premises can play a meaningful role in the evaluation. Second, it is modularly defined to leverage on any GS for QBAFs. We also define a set of novel properties for our GS and study their suitability alongside a set of existing properties (adapted to our setting) for two instantiations of our GS, demonstrating their advantages over existing approaches.

cs.AI↗

arXiv2025

How Do People Revise Inconsistent Beliefs? Examining Belief Revision in Humans with User Studies

Understanding how humans revise their beliefs in light of new information is crucial for developing AI systems which can effectively model, and thus align with, human reasoning. While theoretical belief revision frameworks rely on a set of principles that establish how these operations are performed, empirical evidence from cognitive psychology suggests that people may follow different patterns when presented with conflicting information. In this paper, we present three comprehensive user studies showing that people consistently prefer explanation-based revisions, i.e., those which are guided by explanations, that result in changes to their belief systems that are not necessarily captured by classical belief change theory. Our experiments systematically investigate how people revise their beliefs with explanations for inconsistencies, whether they are provided with them or left to formulate them themselves, demonstrating a robust preference for what may seem non-minimal revisions across different types of scenarios. These findings have implications for AI systems designed to model human reasoning or interact with humans, suggesting that such systems should accommodate explanation-based, potentially non-minimal belief revision operators to better align with human cognitive processes.

cs.AI↗

arXiv2025

On the Utility of Accounting for Human Beliefs about AI Intention in Human-AI Collaboration

To enable effective human-AI collaboration, merely optimizing AI performance without considering human factors is insufficient. Recent research has shown that designing AI agents that take human behavior into account leads to improved performance in human-AI collaboration. However, a limitation of most existing approaches is their assumption that human behavior remains static, regardless of the AI agent's actions. In reality, humans may adjust their actions based on their beliefs about the AI's intentions, specifically, the subtasks they perceive the AI to be attempting to complete based on its behavior. In this paper, we address this limitation by enabling a collaborative AI agent to consider its human partner's beliefs about its intentions, i.e., what the human partner thinks the AI agent is trying to accomplish, and to design its action plan accordingly to facilitate more effective human-AI collaboration. Specifically, we developed a model of human beliefs that captures how humans interpret and reason about their AI partner's intentions. Using this belief model, we created an AI agent that incorporates both human behavior and human beliefs when devising its strategy for interacting with humans. Through extensive real-world human-subject experiments, we demonstrate that our belief model more accurately captures human perceptions of AI intentions. Furthermore, we show that our AI agent, designed to account for human beliefs over its intentions, significantly enhances performance in human-AI collaboration.

cs.AI↗