Search: “multimodal”

arXiv:2604.07737v1 Announce Type: new Abstract: While transformer-based Large Language Models (LLMs) theoretically support massive context windows, they suffer from severe performance degradation when

arXiv:2504.13378v2 Announce Type: replace-cross Abstract: Generating high-quality, photorealistic textures for 3D human avatars remains a fundamental yet challenging task in computer vision and multim

arXiv:2604.07592v1 Announce Type: new Abstract: Embodied-AI agents must reason about how objects move and interact in 3-D space over time, yet existing smaller frontier Large Language Models (LLMs) st

arXiv:2604.06990v1 Announce Type: cross Abstract: Psychological stress is clinically relevant in cardio-oncology, yet it is typically assessed only through patient-reported outcome measures (PROMs) an

arXiv:2412.10437v3 Announce Type: replace Abstract: Generating high-quality Scalable Vector Graphics (SVGs) from text remains a significant challenge. Existing LLM-based models that generate SVG code

arXiv:2604.06375v1 Announce Type: new Abstract: AI-driven symptom analysis systems face persistent challenges in reliability, interpretability, and hallucination. End-to-end generative approaches ofte

arXiv:2604.06274v1 Announce Type: cross Abstract: In recent years, the pace of development of information technology in various areas has increased drastically, forcing cybersecurity specialists to co

arXiv:2604.07634v1 Announce Type: new Abstract: Streaming vision-language models (VLMs) continuously generate responses given an instruction prompt and an online stream of input frames. This is a core

Having a model like Gemma 4, which is perfectly adequate for everyday use in many cases, runs locally, is free, and secure, still feels unreal. We have a very good AI that costs nothing, uses hardly a

Routing every task to your largest model burns tokens, adds latency, and inflates costs. @AI21Labs' Maestro Orchestration Meta Model (OMM) is the layer above your stack that dynamically selects the ri

Common Failure Modes Break VLM-Powered OCR in Production. 🔁 Repetition Loops — model spirals into infinite whitespace, exhausts resources, cascades latency across your system 🛑 Recitation Errors — saf

Google's Gemma 4, launched on April 2, 2026, surpassed 10 million downloads within its first week, according to Google CEO Sundar Pichai, pushing the cumulative total for all Gemma models past 500 ...

NEW: Meta announces Muse Spark. All you need to know: * It's their new multi-modal reasoning model. * Strong at multi-agent orchestration and multi-modal reasoning. * Contemplating mode orchestrates m

Safetensors, a secure tensor serialization format originally developed by Hugging Face, has joined the PyTorch Foundation as a foundation-hosted project under the Linux Foundation, alongside DeepSp...

AlphaXiv introduced GLM-5.1 as the underlying model powering its research paper understanding features on the alphaXiv platform, enabling users to highlight any section of a paper to ask contextual...

arXiv:2610.22065v1 Announce Type: new Abstract: Transformer-based models have become the backbone of a wide range of tasks, but their intensive computation and data movement incur substantial energy o

arXiv:2610.21882v1 Announce Type: cross Abstract: Embodied spatial intelligence requires more than visual recognition: models must localize entities, reason about their relations, and express this kno

arXiv:2610.21534v1 Announce Type: new Abstract: Assist-as-needed (AAN) control must distinguish between changes in intended motion and changes in effective voluntary torque capability, even though bot

arXiv:2610.21011v1 Announce Type: new Abstract: Identifying individual dogs from barks in real-world videos remains challenging and largely unexplored. Solving this problem would enable long-term beha

Beautiful work from @perplexity_ai @AravSrinivas with weights on @huggingface of course! https://huggingface.co/perplexity-ai/pplx-decider-flash-v1.2 https://huggingface.co/perplexity-ai/pplx-decider-

arXiv:2610.21696v1 Announce Type: new Abstract: Markerless intraoperative tracking of spinal anatomy via RGB-D sensors is a promising alternative to invasive bone-anchored fiducials, yet no existing c

arXiv:2610.22187v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used in computational social science to assign scalar scores to social and psychological constructs. How

arXiv:2610.20679v1 Announce Type: new Abstract: Representation-based safety defenses for vision-language models (VLMs) often learn to separate harmful from benign inputs, but separation alone does not

deepagents now supports gemini 4 argon from @GoogleDeepMind! it supports 1M input tokens, long decode (up to 1M), and agentic video understanding! docs: https://docs.langchain.com/oss/python/integrati

arXiv:2610.22160v1 Announce Type: cross Abstract: Test-time context optimization adapts frozen language models through textual feedback, but ordinary conditional generation provides no separate contro

arXiv:2610.21180v1 Announce Type: new Abstract: General-purpose dexterous manipulation requires policies to transfer knowledge from related tasks to new tasks with only a few target demonstrations. Ex

arXiv:2610.21767v1 Announce Type: new Abstract: Can AI agents organize scientific investigations that lead to well-supported conclusions? To evaluate this capability, we introduce PhysClosureBench, a

arXiv:2610.20632v1 Announce Type: new Abstract: Modern healthcare generates increasingly complex medical reports that challenge clinical professionals in terms of time and interpretive burden. Manual

arXiv:2610.21794v1 Announce Type: new Abstract: Despite the remarkable generalization capabilities of Large Language Models, generating clinically accurate and structured radiology reports from medica

arXiv:2610.21007v1 Announce Type: new Abstract: Data selection reallocates training support across visual concepts, including objects, attributes, color, texture, and backgrounds. Different allocation

arXiv:2610.22276v1 Announce Type: new Abstract: Spatial intelligence lets an agent perceive and act in the physical world, and many of its judgments must be made at interaction rates: which spatial re

Muse Spark Open is now live on Wally, day 0 Best US open model thanks to @finkd 370 tok/s, ~2x faster than any provider on OpenRouter. $5 Free on signup. Media 1/ muse spark open is the most capable U

arXiv:2610.22234v1 Announce Type: cross Abstract: Real desktop automation is neither purely visual nor purely programmatic. Agents must understand user-facing interfaces while efficiently manipulating

arXiv:2610.21380v1 Announce Type: new Abstract: Reinforcement learning has brought strong gains to language model reasoning, but its improvements on visual tasks remain less consistent. One difficulty

arXiv:2610.22155v1 Announce Type: new Abstract: Real-time interaction requires agents to perceive evolving environments and act under temporal constraints, yet existing evaluations of real-time stream

arXiv:2610.21539v1 Announce Type: new Abstract: Diffusion models (DMs) have steadily demonstrated outstanding performance in high-fidelity, controllable generation across diverse scenarios, following

arXiv:2610.20766v1 Announce Type: new Abstract: Imagined speech decoding from EEG offers a promising route toward naturalistic brain-computer interfaces. Traditional closed-set classifiers treat seman

arXiv:2610.21913v1 Announce Type: new Abstract: Continuous physical human-robot interaction requires robots to rapidly adapt their control behaviour without interrupting task execution. However, exist

arXiv:2411.16164v2 Announce Type: replace Abstract: Text-to-image synthesis (T2I) focuses on generating high-quality images from textual descriptions In this survey, we review 490 papers on T2I spanni

arXiv:2610.21477v1 Announce Type: new Abstract: Medical calculators translate raw patient observations into clinically actionable values and scores. However, applying these calculators directly to fre

arXiv:2610.20964v1 Announce Type: cross Abstract: Existing long-document QA benchmarks rarely combine large-scale human-authored questions with detailed visual evidence annotations, limiting diagnosis

arXiv:2610.22009v1 Announce Type: new Abstract: Vision-language models (VLMs) are increasingly used to interpret religious, educational, and culturally sensitive visual content, yet their reliability

arXiv:2610.21159v1 Announce Type: new Abstract: On-screen text can describe what a video shows, or it can frame what the video is supposed to mean. Current benchmarks treat both alike, testing whether

arXiv:2610.21915v1 Announce Type: new Abstract: Vision-language-action (VLA) policies are typically trained with action prediction as the primary supervision signal, since robots ultimately execute ac

A big Google strength in AI has always been finding interesting uses for the models they have. They pioneered deep research, podcasting-from-information and all of NotebookLM etc. With Gemini 4, haven

arXiv:2610.15956v1 Announce Type: new Abstract: Vision-language-action models for embodied navigation typically rely on RGB observations, limiting their applicability in lights-out industrial environm