Search: “benchmark”

HAD: Hallucination-Aware Diffusion Priors for 3D Reconstruction

arXiv:2605.16873v1 Announce Type: new Abstract: Diffusion priors have recently demonstrated strong capability in enhancing the quality of sparse-view 3D reconstruction by augmenting training views at

researcharxiv-cs-cv

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Helpful to a Fault: Measuring Illicit Assistance in Multi-Turn, Multilingual LLM Agents

DGX agent

arXiv:2602.16346v3 Announce Type: replace Abstract: LLM-based agents execute real-world workflows via tools and memory. These affordances enable ill-intended adversaries to also use these agents to ca

safetyarxiv-cs-cl

19 May 2026

How Off-Policy Can GRPO Be? Mu-GRPO for Efficient LLM Reinforcement Learning

arXiv:2605.17570v1 Announce Type: cross Abstract: Group Relative Policy Optimization (GRPO) has been a key driver of recent progress in reinforcement learning with verifiable rewards (RLVR) for large

HydroAgent: Closing the Gap Between Frontier LLMs and Human Experts in Hydrologic Model Calibration via Simulator-Grounded RL

model-releasesarxiv-cs-lg

arXiv:2605.17792v1 Announce Type: new Abstract: Calibrating distributed hydrologic models is a critical bottleneck across operational water resources management - streamflow prediction, reservoir oper

Applications

Hypergraph Pattern Machine: Compositional Tokenization for Higher-Order Interactions

arXiv:2605.16527v1 Announce Type: cross Abstract: Hypergraphs model higher-order relations that drive real-world decisions, from drug prescriptions to recommendations. A central structural signal in s

applicationsarxiv-cs-ai

Local Ai

IdGlow: Dynamic Identity Modulation for Multi-Subject Generation

arXiv:2603.00607v2 Announce Type: replace-cross Abstract: Multi-subject image generation requires seamlessly harmonizing multiple reference identities within a coherent scene. However, existing method

local-aiarxiv-cs-ai

Implicit Hierarchical GRPO: Decoupling Tool Invocation from Execution for Tool-Integrated Mathematical Reasoning

arXiv:2605.18500v1 Announce Type: new Abstract: Large language models (LLMs) have increasingly leveraged tool invocation to enhance their reasoning capabilities. However, existing approaches typically

Improving MLLM Training Efficiency via Stage-Aware Sparsity

arXiv:2509.18150v2 Announce Type: replace-cross Abstract: Multimodal Large Language Models (MLLMs) have demonstrated outstanding performance across a variety of domains. However, training MLLMs is oft

safetyarxiv-cs-ai

I/O 2026: Welcome to the agentic Gemini era

At I/O 2026, Google announced that AI is transitioning from something users actively open to a background service that completes tasks automatically. The company introduced Gemini Spark, a new agentic

model-releasesgoogle-ai

Applications

Isolating Nonlinear Independent Sources in fMRI with eta-TCVAE Models

arXiv:2605.16708v1 Announce Type: new Abstract: Learning meaningful latent representations from nonlinear fMRI data remains a fundamental challenge in neuroimaging analysis. Traditional independent co

applicationsarxiv-cs-lg

KairosHope: A Next-Generation Time-Series Foundation Model for Specialized Classification via Dual-Memory Architecture

arXiv:2605.18657v1 Announce Type: cross Abstract: Time Series Foundation Models (TSFMs) have demonstrated notable success in general-purpose forecasting tasks; however, their adaptation to specialized

Kelvin v1.0: A Neural Pre-Encoder for H.264: A standards-compliant learned preprocessor with -27.62% BD-VMAF on UVG

model-releasesarxiv-cs-cv

arXiv:2605.16376v1 Announce Type: cross Abstract: Kelvin is a lightweight learned pre-encoder that sits in front of an unmodified libx264 encoder. It applies content-adaptive pixel adjustments, bounde

KISS - Knowledge Infrastructure for Scientific Simulation: A Scaffolding for Agentic Earth Science

arXiv:2605.17856v1 Announce Type: new Abstract: Process-based simulation models encode decades of scientific understanding across the Earth sciences, yet the communities most exposed to climate risk a

KVCapsule: Efficient Sequential KV Cache Compression for Vision-Language Models with Asymmetric Redundancy

arXiv:2605.16439v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) have emerged as a critical and fast-growing extension of Large Language Models (LLMs) that enable multimodal reasoning t

Hardware

KVDrive: A Holistic Multi-Tier KV Cache Management System for Long-Context LLM Inference

arXiv:2605.18071v1 Announce Type: new Abstract: Supporting long-context LLMs is challenging due to the substantial memory demands of the key-value (KV) cache. Existing offloading systems store the ful

hardwarearxiv-cs-cl

LARGER: Lexically Anchored Repository Graph Exploration and Retrieval

arXiv:2605.16352v1 Announce Type: cross Abstract: Repository-level coding agents must first localize the files and symbols relevant to a task; failures at this stage can cascade across downstream obje

LASAR: Towards Spatio-temporal Reasoning with Latent Cognitive Map

arXiv:2605.16899v1 Announce Type: new Abstract: A fundamental challenge in embodied AI is verifying if agents build internal models of spatial structure or merely learn to mimic task-specific expert t

agentsarxiv-cs-cv

Latent Action Reparameterization for Efficient Agent Inference

arXiv:2605.18597v1 Announce Type: new Abstract: Large language model (LLM) agents often rely on long sequences of low-level textual actions, resulting in large effective decision horizons and high inf

Tutorials

Learning from Historical Activations in Graph Neural Networks

arXiv:2601.01123v2 Announce Type: replace-cross Abstract: Graph Neural Networks (GNNs) have demonstrated remarkable success in various domains such as social networks, molecular chemistry, and more. A

tutorialsarxiv-cs-ai

Learning from Self-Debate: Preparing Reasoning Models for Multi-Agent Debate

arXiv:2601.22297v2 Announce Type: replace Abstract: The reasoning abilities of large language models (LLMs) have been substantially improved by reinforcement learning with verifiable rewards (RLVR). A

agentsarxiv-cs-cl

Learning Higher-Order Structure from Incomplete Spatiotemporal Data: Multi-Scale Hypergraph Laplacians with Neural Refinement

arXiv:2605.17316v1 Announce Type: cross Abstract: Sensor networks increasingly govern modern infrastructure, yet the data they lose are rarely missing in the uniform-random patterns assumed by standar

researcharxiv-cs-ai

Learning in Position-Aware Multinomial Logit Bandits: From Multiplicative to General Position Effects

arXiv:2605.17238v1 Announce Type: new Abstract: We study the dynamic joint assortment selection and positioning problem, where the attraction of each product depends on both its intrinsic appeal and i

researcharxiv-cs-lg

Learning Multi-Timescale Abstractions for Hierarchical Combinatorial Planning

arXiv:2605.17058v1 Announce Type: new Abstract: The combination of exponentially large action spaces, stochastic dynamics, and long-horizon decision-making under limited resources makes Sequential Sto

safetyarxiv-cs-lg

Learning to Reason without External Rewards

arXiv:2505.19590v5 Announce Type: replace-cross Abstract: Training large language models (LLMs) for complex reasoning via Reinforcement Learning with Verifiable Rewards (RLVR) is effective but limited

Learning Transferable Topology Priors for Multi-Agent LLM Collaboration Across Domains

arXiv:2605.17359v1 Announce Type: new Abstract: Large language model (LLM)-based multi-agent systems have shown strong potential for complex reasoning by coordinating specialized agents through struct

LERA: LLM-Enhanced RAG for Ad Auction in Generative Chatbots

arXiv:2605.16474v1 Announce Type: cross Abstract: The integration of advertising auction mechanisms into large language model (LLM)-based chatbots presents a significant opportunity for commercializat

LESSViT: Robust Hyperspectral Representation Learning under Spectral Configuration Shift

model-releasesarxiv-cs-cv

arXiv:2605.18541v1 Announce Type: new Abstract: Modeling hyperspectral imagery (HSI) across different sensors presents a fundamental challenge due to variations in wavelength coverage, band sampling,

Applications

LiPS: Lightweight Panoptic Segmentation for Resource-Constrained Robotics

arXiv:2604.00634v2 Announce Type: replace-cross Abstract: Panoptic segmentation is a key enabler for robotic perception, as it unifies semantic understanding with object-level reasoning. However, the

applicationsarxiv-cs-cv

LISA: Language-guided Interference-aware Spatial-Frequency Attention for Driver Gaze Estimation

arXiv:2605.17287v1 Announce Type: new Abstract: Driver gaze estimation serves as a fundamental metric for evaluating driver attentiveness in modern monitoring systems. Beyond being vulnerable to sudde

researcharxiv-cs-cv

LiteFrame: Efficient Vision Encoders Unlock Frame Scaling in Video LLMs

arXiv:2605.17260v1 Announce Type: new Abstract: The fundamental challenge in scaling Video Large Language Models (Video LLMs) to long-form video lies in managing the explosion of visual-token context

researcharxiv-cs-cv

LLM-Guided Communication for Cooperative Multi-Agent Reinforcement Learning

arXiv:2605.18077v1 Announce Type: new Abstract: Communication is a key component in multi-agent reinforcement learning (MARL) for mitigating partial observability, yet prior approaches often rely on i

LLMForge: Multi-Backend Hardware-Aware Neural Architecture Search with Infinite-Head Attention for Edge Language Models

arXiv:2605.17653v1 Announce Type: cross Abstract: Sub-billion-parameter Transformer language models are increasingly deployed on edge devices, where the privacy, latency, and operating-cost advantages

Local Ai

Locally Coherent Parallel Decoding in Diffusion Language Models

arXiv:2603.20216v2 Announce Type: replace-cross Abstract: Diffusion language models (DLMs) have emerged as a promising alternative to autoregressive (AR) models, offering sub-linear generation latency

local-aiarxiv-cs-ai

Long Context Modeling with Ranked Memory-Augmented Retrieval

arXiv:2503.14800v3 Announce Type: replace-cross Abstract: Effective long-term memory management is crucial for language models handling extended contexts. We introduce the Enhanced Ranked Memory Augme

researcharxiv-cs-ai

Hardware

LongLive-2.0: An NVFP4 Parallel Infrastructure for Long Video Generation

arXiv:2605.18739v1 Announce Type: new Abstract: We present LongLive-2.0, an NVFP4-based parallel infrastructure throughout the full training and inference workflow of long video generation, addressing

hardwarearxiv-cs-cv

LoopQ: Quantization for Recursive Transformers

arXiv:2605.16343v1 Announce Type: cross Abstract: Looped language models (LoopLMs) improve parameter efficiency by recursively reusing Transformer blocks, enabling deeper computation under a fixed mod

LPG: Balancing Efficiency and Policy Reasoning in Latent Policy Guardrails

arXiv:2605.17329v1 Announce Type: cross Abstract: Guardrails are a critical safety layer for modern AI systems, but their operating regime is changing. As LLMs are deployed as customized assistants, s

safetyarxiv-cs-ai

M^2FedAQI: Multimodal Federated Learning for Air Quality Prediction on Heterogeneous Edge Devices

model-releasesarxiv-cs-lg

arXiv:2605.16375v1 Announce Type: new Abstract: Accurate air quality prediction is essential for public health, environmental monitoring, and industrial safety. However, most existing approaches rely

Machine Unlearning for Masked Diffusion Language Models

arXiv:2605.18253v1 Announce Type: cross Abstract: Recent masked diffusion language models (MDLMs), such as LLaDA and Dream, have achieved performance comparable to autoregressive large language models

MADP: A Multi-Agent Pipeline for Sustainable Document Processing with Human-in-the-Loop

arXiv:2605.17159v1 Announce Type: new Abstract: Document processing automation remains a critical challenge in enterprise environments, where traditional manual approaches are labor-intensive and erro

MARS: Technical Report for the CASTLE Challenge at EgoVis 2026

arXiv:2605.18176v1 Announce Type: cross Abstract: This report presents MARS, short for Multimodal Agentic Reasoning with Source selection, our system for the CASTLE Challenge at EgoVis 2026. Participa

MATE: Solving Contextual Markov Decision Processes with Memory of Accumulated Transition Embeddings

arXiv:2605.17431v1 Announce Type: cross Abstract: We propose MATE, a simple yet effective memory architecture for solving Contextual Markov Decision Processes (CMDPs), a family of MDPs parameterized b

MAVEN A Multi-Agent Framework for Multicultural Text-to-Video Generation

arXiv:2605.16716v1 Announce Type: cross Abstract: Text-to-video (T2V) generation has rapidly progressed in visual fidelity, yet its ability to faithfully represent multiple cultures within a single pr

Med-V1: Small Language Models for Zero-shot and Scalable Biomedical Evidence Attribution

arXiv:2603.05308v2 Announce Type: replace-cross Abstract: Assessing whether an article supports an assertion is essential for hallucination detection and claim verification. While large language model

MedMIX: Modality-Internal Expert Fusion for Multimodal Medical Diagnosis

arXiv:2605.16639v1 Announce Type: new Abstract: Multimodal clinical prediction faces three challenges: multiple foundation models (FMs) with complementary strengths per modality, pervasive missing mod

researcharxiv-cs-lg

Membership Inference Attacks on Discrete Diffusion Language Models

arXiv:2605.16445v1 Announce Type: cross Abstract: Masked Diffusion Language Models MDLMs replace autoregressive generation with iterative demasking and their privacy properties are largely unstudied.

MemOCR: Layout-Aware Visual Memory for Efficient Long-Horizon Reasoning

arXiv:2601.21468v5 Announce Type: replace Abstract: Long-horizon agentic reasoning necessitates effectively compressing growing interaction histories into a limited context window. Most existing memor

MemRepair: Hierarchical Memory for Agentic Repository-Level Vulnerability Repair

arXiv:2605.17444v1 Announce Type: cross Abstract: Modern software ecosystems face a rapidly growing number of disclosed vulnerabilities, increasing the need for automated repair techniques that can op