arXiv:2510.18913v5 Announce Type: replace-cross Abstract: Direct Preference Optimization (DPO) is effective but brittle under annotator noise and distribution shift because it operates on hard, pairwise labels and only regularizes log-probability differences. We introduce Anchored Direct Preference Optimization (ADPO), a framework that extends preference learning to soft listwise supervision via reference anchoring. ADPO minimizes KL(q || […]
Trustworthy LLM-Mediated Communication: Evaluating Information Fidelity in LLM as a Communicator (LAAC) Framework in Multiple Application Domains
arXiv:2511.04184v1 Announce Type: cross Abstract: The proliferation of AI-generated content has created an absurd communication theater where senders use LLMs to inflate simple ideas into verbose content, recipients use LLMs to compress them back into summaries, and as a consequence neither party engage with authentic content. LAAC (LLM as a Communicator) proposes a paradigm shift […]
Detecting Silent Failures in Multi-Agentic AI Trajectories
arXiv:2511.04032v1 Announce Type: new Abstract: Multi-Agentic AI systems, powered by large language models (LLMs), are inherently non-deterministic and prone to silent failures such as drift, cycles, and missing details in outputs, which are difficult to detect. We introduce the task of anomaly detection in agentic trajectories to identify these failures and present a dataset curation […]
seqme: a Python library for evaluating biological sequence design
arXiv:2511.04239v1 Announce Type: cross Abstract: Recent advances in computational methods for designing biological sequences have sparked the development of metrics to evaluate these methods performance in terms of the fidelity of the designed sequences to a target distribution and their attainment of desired properties. However, a single software library implementing these metrics was lacking. In […]
Mathematical and Computational Nuclear Oncology: Toward Optimized Radiopharmaceutical Therapy via Digital Twins
arXiv:2511.03755v1 Announce Type: new Abstract: This article presents the general framework of theranostic digital twins (TDTs) in computational nuclear medicine, designed to support clinical decision-making and improve cancer patient prognosis through personalized radiopharmaceutical therapies (RPTs). It outlines potential clinical applications of TDTs and proposes a roadmap for successful implementation. Additionally, the chapter provides a conceptual […]
Efficient Reinforcement Learning from Human Feedback via Bayesian Preference Inference
arXiv:2511.04286v1 Announce Type: cross Abstract: Learning from human preferences is a cornerstone of aligning machine learning models with subjective human judgments. Yet, collecting such preference data is often costly and time-consuming, motivating the need for more efficient learning paradigms. Two established approaches offer complementary advantages: RLHF scales effectively to high-dimensional tasks such as LLM fine-tuning, […]
Why Consciousness Should Explain Physical Phenomena: Toward a Testable Theory
arXiv:2511.04047v1 Announce Type: new Abstract: The reductionist approach commonly employed in scientific methods presupposes that both macro and micro phenomena can be explained by micro-level laws alone. This assumption implies intra-level causal closure, rendering all macro phenomena epiphenomenal. However, the integrative nature of consciousness suggests that it is a macro phenomenon. To ensure scientific testability […]
Federated Learning with Gramian Angular Fields for Privacy-Preserving ECG Classification on Heterogeneous IoT Devices
arXiv:2511.03753v1 Announce Type: cross Abstract: This study presents a federated learning (FL) framework for privacy-preserving electrocardiogram (ECG) classification in Internet of Things (IoT) healthcare environments. By transforming 1D ECG signals into 2D Gramian Angular Field (GAF) images, the proposed approach enables efficient feature extraction through Convolutional Neural Networks (CNNs) while ensuring that sensitive medical data […]
Speed at the Cost of Quality? The Impact of LLM Agent Assistance on Software Development
arXiv:2511.04427v1 Announce Type: cross Abstract: Large language models (LLMs) have demonstrated the promise to revolutionize the field of software engineering. Among other things, LLM agents are rapidly gaining momentum in their application to software development, with practitioners claiming a multifold productivity increase after adoption. Yet, empirical evidence is lacking around these claims. In this paper, […]
Leveraging LLM-based agents for social science research: insights from citation network simulations
arXiv:2511.03758v1 Announce Type: cross Abstract: The emergence of Large Language Models (LLMs) demonstrates their potential to encapsulate the logic and patterns inherent in human behavior simulation by leveraging extensive web data pre-training. However, the boundaries of LLM capabilities in social simulation remain unclear. To further explore the social attributes of LLMs, we introduce the CiteAgent […]