6056 papers
This paper introduces GFM-Retriever, a novel GraphRAG framework that leverages a pre-trained Graph Foundation Model for cross-domain subgraph retrieval and an Information Bottleneck-based selector to extract minimal, sufficient reasoning paths, thereby achieving state-of-the-art performance in multi-hop question answering without relying on domain-specific heuristics.
The paper introduces FreeFly-Thinking, an end-to-end Vision-Language Navigation framework for UAVs that leverages a two-stage training strategy and explicit chain-of-thought reasoning to achieve robust and efficient navigation in complex outdoor urban environments.
This paper addresses the limitation of existing creative activity tracing methods that capture state changes without preserving intent or higher-level structure by proposing three complementary domain-specific approaches: a node-based interface for GenAI, a vocabulary of visual cues for visualization authoring, and a semantic history-embedded programming model.
This mixed-methods study of 37 undergraduate students at a Ghanaian university reveals a strong preference for familiar instant messaging platforms like WhatsApp and Telegram over institutional learning management systems, driven by factors such as convenience, accessibility, and real-time interaction, thereby highlighting the need for educational strategies to align with students' existing digital habits.
This paper proposes the Layered Governance Architecture (LGA), a four-layer framework designed to systematically mitigate execution-layer vulnerabilities in autonomous agent systems, and validates its effectiveness through a bilingual benchmark demonstrating high interception rates of malicious tool calls with minimal latency.
FastSTAR is a training-free acceleration framework for Spacetime Autoregressive (STAR) video generation that utilizes a novel Spatiotemporal Token Pruning strategy to identify and skip redundant computations, achieving up to a 2.01x speedup with minimal quality degradation.
The paper introduces Re, a reinforcement learning framework that enables large language models to dynamically abandon unproductive reasoning paths and restart their solution process, thereby significantly improving reasoning efficiency and performance without requiring preliminary supervised fine-tuning.
This paper proposes a fully onboard, vision-guided optimal control framework that enables agile drone racing through arbitrary gate poses by integrating a novel neural signed distance field (Gate-SDF) with a Model Predictive Path Integral (MPPI) controller to achieve robust, reference-free navigation without relying on explicit pose estimation or precomputed trajectories.
This paper proposes and evaluates an on-premises collaborative framework utilizing multiple large language models with majority voting to efficiently and privately identify cryptographically relevant software packages, thereby addressing the challenges of manual inventory and static analysis limitations in the transition to post-quantum cryptography.
This study utilizes PLS-SEM analysis of data from Ghanaian universities to demonstrate that subjective norms, deterrence, and involvement mechanisms—particularly knowledge sharing—significantly shape contingent employees' attitudes toward information security policies, thereby driving their compliance intentions.
This paper introduces a miniature brain transformer architecture that demonstrates a novel, falsifiable prediction: functional lateralization of hippocampal banks requires the synergistic interaction of a prefrontal working-memory buffer (acting as a symmetry-breaker) and inhibitory callosal coupling, a mechanism that triggers a sharp phase transition in memory performance while a cerebellar fast-path merely accelerates convergence.
VINO is a self-supervised learning framework that overcomes the "co-occurrence trap" in dense video by using a teacher-student distillation approach with structural priors to force representations to focus on foreground objects rather than background context, achieving state-of-the-art unsupervised object discovery performance.
This paper proposes a hybrid Learning-to-Rank recommender system that leverages deep learning and social context embeddings from Stack Overflow to effectively recommend software bug solutions to developers, achieving nearly 78% accuracy in the top 10 results.
This paper introduces BATDiff, an unsupervised single-image super-resolution model that leverages bivariate A-trous wavelet transforms and cross-scale parent-child dependencies to generate sharper, more structurally consistent high-frequency details while minimizing artifacts and dataset-driven hallucinations.
This paper addresses the unexplored challenge of avoiding redundancy in data lakes by defining the Novel Table Search (NTS) problem, proposing an efficient penalization-based approximation algorithm called ANTs, and demonstrating through extensive experiments that it outperforms existing methods in capturing syntactic novelty while maintaining low execution time.
The paper proposes HY-WU, a memory-first adaptation framework that replaces static weight overwriting with a functional neural memory module to synthesize instance-specific weight updates on-the-fly, thereby enabling robust continual learning and personalization without degrading previously learned behaviors.
This paper proposes a soft actor-critic-based reinforcement learning framework for vehicle-to-grid voltage regulation that effectively coordinates single and multi-hub charging systems while prioritizing battery health and fleet availability, demonstrating robust performance comparable to standard droop controllers under both nominal and aggressive overloading conditions.
FabricGen is an end-to-end framework that generates realistic woven fabric materials from text by decoupling macro-scale texture synthesis via fine-tuned diffusion models from micro-scale weaving pattern generation using a specialized LLM-driven procedural geometric model.
This paper introduces PresentBench, a fine-grained, rubric-based benchmark comprising 238 instances and detailed binary checklists that enables more reliable, human-aligned evaluation of automated slide generation, revealing that NotebookLM currently outperforms other methods in this domain.
This paper introduces LEPA, a learned architecture that conditions on geometric augmentations to accurately predict transformed satellite image embeddings, effectively overcoming the limitations of standard interpolation in non-convex geospatial foundation model manifolds and significantly improving geometric adjustment performance.