AI Research news and analysis on Web Pulse: 22244 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
A new arXiv paper (2609.13672v1) introduces recoverability as a system primitive for long-horizon AI agents, making the reuse of saved state an explicit decision rather than an automatic continuation. The paper's referen…
The FedV-KGQA poster paper reports that federated fusion across vertically partitioned knowledge graphs recovers most of the accuracy of a centralized system, while a single silo recovers little. The work, presented as a…
Researchers proposed GeoSkill, an experience-driven hierarchical skill learning framework for geospatial agents, detailed in arXiv paper 2609.13667v1. GeoSkill combines a Hierarchical Skill Bank, split into a Planning Sk…
Researchers introduced MACE, a Multi-Agent Candidate Event acquisition method that uses evidence-specialized LLM agents to refine event structure before linking, according to a new arXiv paper (arXiv:2609.13670v1). MACE …
Researchers proposed a Position Encoding-Based Deformable Spatial Aggregation Module (PEBDSAM) that addresses over-smoothing, over-compression, fixed receptive fields, and topological noise in graph neural networks, acco…
A new arXiv survey (2609.13731v1) synthesizes 206 foundational studies and regulatory standards into a systems-security reference framework for trustworthy agentic AI, formalizing the general agent architecture as a stat…
Researchers introduced JaxAHT, described as the first open-source, JAX-based library for the Ad Hoc Teamwork research lifecycle, achieving approximately 95x wall-clock speedup over PyTorch counterparts. JaxAHT provides a…
A physiology-guided hybrid framework combining a convolutional neural network branch with a morphology-prior branch estimated blood pressure from photoplethysmography (PPG) with mean absolute errors of 3.77 mmHg for syst…
Researchers introduced Counterfactual Search for Grounding Regions (CSGR), a scalable pipeline that labels answer-critical image regions in visual question answering (VQA) data by perturbing candidate regions and measuri…
A three-stage pipeline built from Stable Diffusion 1.5 components generated synthetic leprosy lesion images whose internal perceptual diversity (0.662) was statistically indistinguishable from that of the real leprosy se…
Researchers introduced Abstract-LoRA, a single-image style transfer method that trains lightweight LoRA adapters on specific U-Net blocks in diffusion models, according to an arXiv paper (arXiv:2609.13239v1). Building on…
Researchers Alpay Ozkan and colleagues introduced a pixel-wise planarity prediction framework for monocular plane segmentation from a single RGB image, published as arXiv:2609.13246v1. The method adds a dedicated planari…
ArtSociety, a multi-agent framework that coordinates heterogeneous multimodal experts with two training-free controllers, achieved an Overall Score of 0.8870 on the official AffectiveArt Multidimensional Art Emotion Unde…
A new arXiv paper (2609.13251v1) proposes a subject clarity outpainting framework that combines vision-language model (VLM)-guided semantic conditioning with multiscale wavelet supervision to preserve subject detail when…
A new arXiv survey (arXiv:2609.13255v1) reviews multimodal facial state analysis, covering core tasks, representative methods, and datasets for facial expression recognition, action unit (AU) detection, and face-based so…
A structured temporal video reasoning pipeline combining a discrete EventGraph, a continuous EventField, and a human-readable EventGlyph view achieved 0.98 overall accuracy on a calibrated EPIC-KITCHENS subset of 10 vide…
A new arXiv paper (2609.13262v1) proposes a holistic framework for 3D Gaussian Splatting scene extrapolation and interpolation that replaces the error-prone generate-reconstruct-shift cycle with an independent camera vie…
Researchers introduced TryOnReward, a fine-grained reward model for virtual try-on (VTON) built on a vision-language backbone that uses a foveation calibration objective and jointly optimizes pairwise preferences and per…
A new arXiv paper (2609.13422v1) introduces Vibe Patenting, an end-to-end patent-drafting testbed showing that a separately-invoked LLM judge providing structured feedback consistently improves judge-assessed patent draf…
Researchers introduced OrchSLM, a routing framework that unifies existing non-interactive small language model (SLM) orchestration methods and exposes their design choices as controllable parameters, according to the arX…