1062584 results (page 57 of 42504)
-
Detecting Hallucinations in SpeechLLMs at Inference Time Using Attention Maps
Hallucinations in Speech Large Language Models (SpeechLLMs) pose significant risks, yet existing detection methods typically rely on gold-standard outputs that are costly or impractical to obtain. Moreover, hallucination detection methods developed for text-based LLMs do not directly capture audio-specific signals. We investigate four attention-derived metrics: AUDIORATIO, AUDIOCONSISTENCY, AUDIOE…
-
EgoSelf: From Memory to Personalized Egocentric Assistant
Egocentric assistants often rely on first-person view data to capture user behavior and context for personalized services. Since different users exhibit distinct habits, preferences, and routines, such personalization is essential for truly effective assistance. However, effectively integrating long-term user data for personalization remains a key challenge. To address this, we introduce EgoSelf, …
-
Information-to-energy trade-offs and the optimal alphabet of polymer replication
We analyze information transmission in a recently proposed coarse-grained model of polymer replication by framing it as a communication channel between templates and copies. By calculating the mutual information in the steady-state limit of long chains, we recover the accurate-random phase diagram and establish that the information per-monomer depends solely on template specificity within the accu…
-
Structure-guided molecular design with contrastive 3D protein-ligand learning
Structure-based drug discovery faces the dual challenge of accurately capturing 3D protein-ligand interactions while navigating ultra-large chemical spaces to identify synthetically accessible candidates. In this work, we present a unified framework that addresses these challenges by combining contrastive 3D structure encoding with autoregressive molecular generation conditioned on commercial comp…
-
Detecting Data Contamination in Large Language Models
Large Language Models (LLMs) utilize large amounts of data for their training, some of which may come from copyrighted sources. Membership Inference Attacks (MIA) aim to detect those documents and whether they have been included in the training corpora of the LLMs. The black-box MIAs require a significant amount of data manipulation; therefore, their comparison is often challenging. We study state…
-
Separating Geometry from Probability in the Analysis of Generalization
The goal of machine learning is to find models that minimize prediction error on data that has not yet been seen. Its operational paradigm assumes access to a dataset $S$ and articulates a scheme for evaluating how well a given model performs on an arbitrary sample. The sample can be $S$ (in which case we speak of ``in-sample'' performance) or some entirely new $S'$ (in which case we speak of ``ou…
-
Enhancing Construction Worker Safety in Extreme Heat: A Machine Learning Approach Utilizing Wearable Technology for Predictive Health Analytics
Construction workers are highly vulnerable to heat stress, yet tools that translate real-time physiological data into actionable safety intelligence remain scarce. This study addresses this gap by developing and evaluating deep learning models, specifically a baseline Long Short-Term Memory (LSTM) network and an attention-based LSTM, to predict heat stress among 19 workers in Saudi Arabia. Using G…
-
On Reasoning-Centric LLM-based Automated Theorem Proving
Automated theorem proving is fundamental to formal methods, and the recent trend is to integrate large language models (LLMs) and proof assistants to form effective proof agents. While existing proof agents show promising performance, they inadequately leverage reasoning capabilities of modern LLMs in high-level planning and self-critique. We argue that proof agents should not merely generate tact…
-
Understanding supernova gravitational waves with protoneutron star asteroseismology
Supernovae are one of the most promising gravitational wave sources. But, since the system of the supernovae is nearly spherically symmetric, the expected gravitational waves from them are relatively weak, compared to the case of the compact binary mergers. Thus, at least using the current gravitational wave detectors, only the gravitational waves from a supernova that occurred in our galaxy could…
-
Paparazzo: Active Mapping of Moving 3D Objects
Current 3D mapping pipelines generally assume static environments, which limits their ability to accurately capture and reconstruct moving objects. To address this limitation, we introduce the novel task of active mapping of moving objects, in which a mapping agent must plan its trajectory while compensating for the object's motion. Our approach, Paparazzo, provides a learning-free solution that r…
-
Local power of approximation in hierarchical spline spaces on weakly admissible meshes
We study local approximation properties in hierarchical spline spaces through a twofold approach. First, we design and analyze a robust adaptive refinement algorithm to construct locally graded meshes. Second, we establish rigorous stability and approximation results using computationally efficient quasi-interpolation operators. The primary contribution is the analysis of weakly admissible hierarc…
-
Characterizing and spectrally modeling embedded FUor eruptions in the near-infrared
Context. Episodic accretion in young stellar objects (YSOs) is thought to play a critical role in addressing the "luminosity problem" associated with star formation. However, optical surveys tend to bias against sources that are heavily obscured. Infrared time-domain surveys, such as unTimely WISE, facilitate the identification of such sources within the dense star formation regions of our Galaxy.…
-
LoopCTR: Unlocking the Loop Scaling Power for Click-Through Rate Prediction
Scaling Transformer-based click-through rate (CTR) models by stacking more parameters brings growing computational and storage overhead, creating a widening gap between scaling ambitions and the stringent industrial deployment constraints. We propose LoopCTR, which introduces a loop scaling paradigm that increases training-time computation through recursive reuse of shared model layers, decoupling…
-
Taming Actor-Observer Asymmetry in Agents via Dialectical Alignment
Large Language Model agents have rapidly evolved from static text generators into dynamic systems capable of executing complex autonomous workflows. To enhance reliability, multi-agent frameworks assigning specialized roles are increasingly adopted to enable self-reflection and mutual auditing. While such role-playing effectively leverages domain expert knowledge, we find it simultaneously induces…
-
Emotion-Cause Pair Extraction in Conversations via Semantic Decoupling and Graph Alignment
Emotion-Cause Pair Extraction in Conversations (ECPEC) aims to identify the set of causal relations between emotion utterances and their triggering causes within a dialogue. Most existing approaches formulate ECPEC as an independent pairwise classification task, overlooking the distinct semantics of emotion diffusion and cause explanation, and failing to capture globally consistent many-to-many co…
-
Viability of Big Bang Nucleosynthesis in Some Generalized Horizon Entropies
In this work, we investigate the viability of some cosmological models derived from generalized horizon entropies, using Big Bang Nucleosynthesis (BBN) constraints. By analyzing the deviations in the expansion rate, we derive bounds on the model parameters from freeze-out temperature, helium, and deuterium abundances. Our results show that the freeze-out condition provides the most stringent const…
-
DT2IT-MRM: Debiased Preference Construction and Iterative Training for Multimodal Reward Modeling
Multimodal reward models (MRMs) play a crucial role in aligning Multimodal Large Language Models (MLLMs) with human preferences. Training a good MRM requires high-quality multimodal preference data. However, existing preference datasets face three key challenges: lack of granularity in preference strength, textual style bias, and unreliable preference signals. Besides, existing open-source multimo…
-
Dynamically cold discs in high-redshift galaxies: comparison between ALMA observations and TNG50
Observations of highly rotationally supported gas discs in high redshift ($z$ > 3) star-forming galaxies challenge our understanding of galaxy formation, as the prevailing view holds that galaxies in the early universe are dynamically hot due to frequent mergers, gas accretion, and strong stellar feedback. We examined the kinematic properties of massive ($M_{\star} \geq 10^{10}\,M_{\odot}$) star-f…
-
FOCAL: Filtered On-device Continuous Activity Logging for Efficient Personal Desktop Summarization
Desktop interaction streams provide a continuous, privacy-sensitive record of interleaved user tasks. Transforming these streams into task-organized personal logs on-device faces two main challenges: exhaustive Vision-Language Model (VLM) processing strains local resources, and global stream processing causes cross-task context pollution. We present FOCAL (Filtered On-device Continuous Activity Lo…
-
Mesh Memory Protocol: Semantic Infrastructure for Multi-Agent LLM Systems
Teams of LLM agents increasingly collaborate on tasks spanning days or weeks: multi-day data-generation sprints where generator, reviewer, and auditor agents coordinate in real time on overlapping batches; specialists carrying findings forward across session restarts; product decisions compounding over many review rounds. This requires agents to share, evaluate, and combine each other's cognitive …
-
Predicting Scale-Up of Metal-Organic Framework Syntheses with Large Language Models
Scalable synthesis remains the gate between MOF discovery and industrial deployment, as scale-up know-how is fragmented across disparate reports. We introduce ESU-MOF, a literature-mined dataset and a positive-unlabeled learning strategy that fine-tunes large language models to predict scalability potential with 91.4% accuracy, enabling rapid data-driven triage for industrial MOF discovery.
-
Integrating Anomaly Detection into Agentic AI for Proactive Risk Management in Human Activity
Agentic AI, with goal-directed, proactive, and autonomous decision-making capabilities, offers a compelling opportunity to address movement-related risks in human activity, including the persistent hazard of falls among elderly populations. Despite numerous approaches to fall mitigation through fall prediction and detection, existing systems have not yet functioned as universal solutions across ca…
-
InvestChat: Exploring Multimodal Interaction via Natural Language, Touch, and Pen in an Investment Dashboard
We designed and implemented InvestChat, a multimodal tablet-based application that supports stock market exploration with multiple coordinated views and an LLM-powered chat. We evaluated the application with 12 novice investors. Our findings suggest that combining natural language, touch, and pen input during stock market exploration facilitates user engagement. Participants leveraged the modaliti…
-
LiveVLN: Breaking the Stop-and-Go Loop in Vision-Language Navigation
Recent navigation systems achieve strong benchmark results, yet real-world deployment often remains visibly stop-and-go. This bottleneck arises because the sense-inference-execution loop is still blocking: after each new observation, the controller must wait for sensing, transmission, and inference before motion can continue. Reducing action-generation cost alone therefore does not remove redundan…
-
The swept-back multipolar magnetic field of neutron stars: Application to NICER MSP J0030+0451
NICER observations of millisecond pulsars (MSPs) suggest that non-dipolar magnetic fields are required to explain their surface X-ray hotspots. C. Kalapotharakos et al. (2021) modeled the NICER light curve of MSP J0030+0451 (J0030) using a static vacuum offset dipole-plus-quadrupole field and corresponding force-free (FF) solutions to jointly reproduce the X-ray and Fermi-LAT $γ$-ray emission. We …