Search: “inference”

What Should Agents Say? Action-state Communication for Efficient Multi-Agent Systems

model-releasesarxiv-cs-ai

arXiv:2606.05304v1 Announce Type: new Abstract: Multi-agent systems (MAS) built on large language models are typically organized around roles, pipelines, and turn schedules, while the content that age

6 Jun 2026

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

When Good Enough Is Optimal: Multiplication-Only Matrix Inversion Approximation for Quantized Gated DeltaNet

DGX agent

arXiv:2606.06034v1 Announce Type: cross Abstract: Matrix inversion in chunk-wise parallel linear attention is a major bottleneck for long-context modeling, particularly on NPUs, where forward-substitu

researcharxiv-cs-ai

6 Jun 2026

Your margin is my opportunity: AI version… The biggest surprise of 2026 is that the capability gap between the best open-weight/source model…

model-releasesclem-delangue--x

Your margin is my opportunity: AI version… The biggest surprise of 2026 is that the capability gap between the best open-weight/source models and the best closed models has narrowed much faster than t

6 Jun 2026

A research team that includes Huawei says it successfully used Huawei's Ascend 910C chips for DeepSeek V4 Pro model's post-training, amid increased US sanctions (Coco Feng/South China Morning Post)

Coco Feng / South China Morning Post: A research team that includes Huawei says it successfully used Huawei's Ascend 910C chips for DeepSeek V4 Pro model's post-training, amid increased US sanctions —

model-releasestechmeme

Activation-Informed Pareto-Guided Low-Rank Compression for Efficient LLM/VLM

arXiv:2510.05544v2 Announce Type: replace Abstract: Large language models (LLM) and vision-language models (VLM) have achieved state-of-the-art performance, but they impose significant memory and comp

Agents

Active Video Perception: Iterative Evidence Seeking for Agentic Long Video Understanding

arXiv:2512.05774v2 Announce Type: replace-cross Abstract: Long video understanding (LVU) is challenging because answering real-world queries often depends on sparse, temporally dispersed cues buried i

agentsarxiv-cs-cl

AdaPlanBench: Evaluating Adaptive Planning in Large Language Model Agents under World and User Constraints

arXiv:2606.05622v1 Announce Type: new Abstract: Planning for real-world problems by language models often involves both world and user constraints, which may not be fully specified upfront and are pro

Aligning Tree-Search Policies with Fixed Token Budgets in Test-Time Scaling of LLMs

arXiv:2602.09574v2 Announce Type: replace Abstract: Tree-search decoding is an effective form of test-time scaling for large language models (LLMs), but real-world deployment often imposes a fixed per

Ask-to-Clarify: Resolving Instruction Ambiguity through Multi-turn Dialogue

arXiv:2509.15061v2 Announce Type: cross Abstract: The ultimate goal of embodied agents is to create collaborators that can interact with humans, not mere executors that passively follow instructions.

applicationsarxiv-cs-cv

AURA: Intent-Directed Probing for Implicit-Need Surfacing in Situated LLM Agents

arXiv:2606.05557v1 Announce Type: new Abstract: A situated query like 'where is Lin Wei?' often encodes more than its literal content: the user may also want to know whether Lin Wei is free, in a good

b9528

local-aillama-cpp-releases

Release b9528 of llama.cpp was published on June 5, 2026. This build includes a UI update to run npm install when package-lock.json is newer than node_modules, and provides pre-compiled binaries for m

Beyond tokens: a unified framework for latent communication in LLM-based multi-agent systems

arXiv:2606.05711v1 Announce Type: new Abstract: Multi-agent systems built on large language models (LLMs) have become a prevailing paradigm for tackling complex reasoning, planning, and tool-use tasks

safetyarxiv-cs-cl

BMCR: Adaptive Backbone Module Composition via Reinforcement Learning for Remote Sensing Object Detection

arXiv:2606.05586v1 Announce Type: new Abstract: In remote sensing object detection, Convolutional Neural Networks (CNNs) excel at capturing local details while Vision Transformers (ViTs) are better at

Code2LoRA: Hypernetwork-Generated Adapters for Code Language Models under Software Evolution

arXiv:2606.06492v1 Announce Type: cross Abstract: Code language models need repository-level context to resolve imports, APIs, and project conventions. Existing methods inject this knowledge as long i

Comparison of Deep Learning Frameworks For Rice Disease Mapping From UAV Multispectral Imaging

arXiv:2606.06359v1 Announce Type: new Abstract: In this study, UAV multispectral imagery is used to segment the severity of bacterial leaf blight (BLB) in rice using convolutional neural networks (CNN

Complexity-Balanced Diffusion Splitting

arXiv:2606.06477v1 Announce Type: new Abstract: Standard continuous-time generative models rely on monolithic architectures that must navigate vastly different signal regimes, from isotropic noise to

Correcting Prompt Dependence in LLM Benchmarks: A Bayesian Hierarchical Model with Embedding-Space Clustering

arXiv:2510.05709v2 Announce Type: replace-cross Abstract: LLM benchmarking metrics often misstate performance and uncertainty as they rely on two assumptions that frequently do not hold in practice: (

Cosine Misleads: Auxiliary Losses Reshape Vision Language Models, Not Their Latents

arXiv:2606.05753v1 Announce Type: new Abstract: Latent visual reasoning (LVR) inserts supervised latent tokens between perception and answer generation in vision-language models (VLMs). The field uses

safetyarxiv-cs-cv

Do Models Share Safety Representations? Cross-Model Steering for Safe Visual Generation

arXiv:2606.05290v1 Announce Type: new Abstract: Recent progress in generative modeling has made safety control a central challenge, yet existing approaches remain largely model-specific, requiring ret

safetyarxiv-cs-cv

DocHop-QA: Towards Multi-Hop Reasoning over Multimodal Document Collections

arXiv:2508.15851v2 Announce Type: replace Abstract: Despite rapid progress in large language models (LLMs), current QA benchmarks still overlook the core challenge of real-world scientific information

Drive-KD: Multi-Teacher Distillation for VLMs in Autonomous Driving

model-releasesarxiv-cs-cv

arXiv:2601.21288v2 Announce Type: replace-cross Abstract: Autonomous driving is an important and safety-critical task, and recent advances in LLMs/VLMs have opened new possibilities for reasoning and

Dynamic Thinking-Token Selection for Efficient Reasoning in Large Reasoning Models

arXiv:2601.18383v2 Announce Type: replace-cross Abstract: Large Reasoning Models (LRMs) excel at solving complex problems by explicitly generating a reasoning trace before deriving the final answer. H

Efficient Punctuation Restoration via Weighted Lookahead Scoring Method for Streaming ASR Systems

arXiv:2606.05179v1 Announce Type: new Abstract: Punctuation restoration improves ASR (Automatic Speech Recognition) readability. However streaming ASR requires online decisions with limited future con

EgoAction: Egocentric Action Composition with Reliability-Aware Temporal Fusion for the EPIC-KITCHENS Action Detection Challenge at CVPR 2026

arXiv:2605.24496v2 Announce Type: replace Abstract: The EPIC-KITCHENS-100 Action Detection challenge evaluates whether a model can localize the start and end of each action in long untrimmed egocentri

EVE: A Generator-Verifier System for Generative Policies

arXiv:2512.21430v2 Announce Type: replace Abstract: Visuomotor policies based on generative such as diffusion and flow-matching have shown strong performance for robotics applications but degrade unde

safetyarxiv-cs-ro

Agents

Executable Schema Contracts: From Automatic Ingestion to Multi-Source Retrieval

arXiv:2606.05415v1 Announce Type: new Abstract: Real-world data spans tables, documents, and semi-structured files with implicit semantics. Querying this data requires integrating evidence across inco

agentsarxiv-cs-cl

GMBFormer: An NDVI-Guided Global Memory Bank Transformer for Urban Green-Space Extraction from Ultra-High-Resolution Imagery

arXiv:2606.06363v1 Announce Type: new Abstract: Urban green-space extraction from ultra-high-resolution (UHR) imagery is commonly performed patch by patch, which limits semantic reuse among spatially

Tools

Highlights: 👉 Design-first generation for ads, posters, packaging, and product visuals 👉 Strong typography and multilingual text rendering…

Highlights: 👉 Design-first generation for ads, posters, packaging, and product visuals 👉 Strong typography and multilingual text rendering 👉 Precise layout and color-palette control for brand workflow

toolstogether-ai--x

Humans' ALMANAC: A Human Collaboration Dataset of Action-Level Mental Model Annotations for Agent Collaboration

arXiv:2606.06388v1 Announce Type: cross Abstract: Recent advances in LLM agents have enabled complex cognitive capabilities, such as multi-step reasoning, planning, and tool use, that increasingly pos

HyperVis: Continuous Latent Visual Relational Graphs on the Lorentz Hyperboloid for Compositional Reasoning

arXiv:2606.06100v1 Announce Type: new Abstract: Vision-Language Models (VLMs) struggle with compositional reasoning that requires understanding inter-object relationships. A natural remedy is to injec

Improving Heart-Focused Medical Question Answering in LLMs via Variance-Aware Rubric Rewards with GRPO

arXiv:2606.05174v1 Announce Type: new Abstract: Large Language Models (LLMs) have shown strong promise in healthcare applications. Yet deploying general-purpose models in real-world settings remains d

local-aiarxiv-cs-cl

In-Context Multiple Instance Learning

arXiv:2606.06458v1 Announce Type: cross Abstract: Multiple Instance Learning (MIL) addresses problems where supervision is available at the level of bags of instances and has been successfully applied

applicationsarxiv-cs-cv

Introducing Ideogram 4 from @ideogram_ai on Together AI, an open image model built for design with strong text rendering, layout control, an…

applicationstogether-ai--x

Introducing Ideogram 4 from @ideogram_ai on Together AI, an open image model built for design with strong text rendering, layout control, and native 2K image generation. AI natives can now use Ideogra

IR3DE: A Linear Router for Large Language Models

arXiv:2606.06098v1 Announce Type: new Abstract: Foundational Large Language Models (LLMs) demonstrate proficiency on a wide range of general tasks, and achieve remarkable results on various specialize

LANTERN: Layered Archival and Temporal Episodic Retrieval Network for Long-Context LLM Conversations

arXiv:2606.05182v1 Announce Type: new Abstract: Large language models discard critical details when conversation history is compacted to fit within finite context windows. We present LANTERN (Layered

applicationsarxiv-cs-cl

Less is MoE: Trimming Experts in Domain-Specialist Language Models

arXiv:2606.05538v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) models achieve strong performance through conditional computation, but their large parameter footprint poses deployment chall

LLM-Enhanced Dialogue Management for Full-Duplex Spoken Dialogue Systems

arXiv:2502.14145v3 Announce Type: replace Abstract: Achieving full-duplex communication in spoken dialogue systems (SDS) requires real-time coordination between listening, speaking, and thinking. This

LoomVideo: Unifying Multimodal Inputs into Video Generation and Editing

model-releasesarxiv-cs-cv

arXiv:2606.06042v1 Announce Type: new Abstract: Developing unified video generation and editing models capable of interpreting interleaved multimodal inputs is a promising yet challenging frontier fie

Mechanistic Insights into Functional Sparsity in Multimodal LLMs via CoRe Heads

arXiv:2606.05843v1 Announce Type: new Abstract: While Multimodal Large Language Models (MLLMs) demonstrate remarkable proficiency on complex vision-language tasks, the mechanisms by which they extract

local-aiarxiv-cs-cl

Membrane: A Self-Evolving Contrastive Safety Memory for LLM Agent Defense

arXiv:2606.05743v1 Announce Type: cross Abstract: Despite advances in safety alignment, large language models remain vulnerable to continuously evolving jailbreaks. Existing fine-tuned safety classifi

safetyarxiv-cs-cl

MIRAI: Prediction and Generation of High-Impact Academic Research

arXiv:2606.05443v1 Announce Type: cross Abstract: The rapid pace of scientific publishing has made the identification and synthesis of high-impact work an increasingly urgent challenge. We introduce M

Multilingual Detection of Alzheimer's Disease from Speech: A Cross-Linguistic Transfer Learning Approach

arXiv:2606.05545v1 Announce Type: new Abstract: The development of multilingual Alzheimer's Disease Dementia (AD) detection models presents significant challenges due to the resource-intensive and tim

NIV: Neural Axis Variations for Variable Font Generation

arXiv:2606.05261v1 Announce Type: new Abstract: Variable fonts enable continuous variation of glyph geometry along semantic design axes such as weight, width, slant, and optical size. However, constru

On Advantage Estimates for Max@K Policy Gradients

arXiv:2606.06080v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards is widely used for post-training reasoning models, but sparse outcome rewards make exploration difficul

safetyarxiv-cs-cl

Parallel Jacobi Decoding for Fast Autoregressive Image Generation

arXiv:2606.05703v1 Announce Type: new Abstract: Autoregressive (AR) models have demonstrated remarkable performance in generating high-fidelity images. However, their inherently sequential next-token

Agents

PathWISE: Multi-Agent Cancer Pathway Triaging Ontology Learning from Clinical Flowcharts

arXiv:2605.25970v2 Announce Type: replace Abstract: Clinical pathways are disseminated as visual flowcharts where spatial topology, arrow direction, colour coding, and font weight encode critical tria

agentsarxiv-cs-cv

PiL-World: A Chunk-Wise World Model for VLA Policy-in-the-Loop Evaluation

arXiv:2606.05773v1 Announce Type: new Abstract: Vision-language-action (VLA) policies operate in a closed loop in real-world robot tasks: a robot observes the scene, executes an action chunk, and cond

safetyarxiv-cs-ro

RadiusFPS: Efficient Farthest Point Sampling on CPUs and GPUs via Spherical Voxel Pruning

arXiv:2606.06255v1 Announce Type: cross Abstract: Point clouds are a primary sensory representation for robotic perception, underpinning LiDAR-based autonomous driving, simultaneous localization and m