All entries

Latent Reasoning with Normalizing Flows

arXiv:2606.06447v1 Announce Type: new Abstract: Large language models often improve reasoning by generating explicit chain-of-thought (CoT), demonstrating the importance of intermediate computation. H

safetyarxiv-cs-cl

LatentSkill: From In-Context Textual Skills to In-Weight Latent Skills for LLM Agents

arXiv:2606.06087v1 Announce Type: new Abstract: Agent systems increasingly use textual skills to encode reusable task procedures, but injecting these skills into the prompt at every step incurs substa

model-releasesarxiv-cs-cl

5 Jun 2026

Local Ai

LeanMarathon: Toward Reliable AI Co-Mathematicians through Long-Horizon Lean Autoformalization

arXiv:2606.05400v1 Announce Type: cross Abstract: Long-horizon autoformalization of research mathematics fails not only at hard lemmas, but at scale: statements drift, dependencies tangle, context dec

local-aiarxiv-cs-cl

Learning Contact Representation for Leg Odometry

arXiv:2606.05501v1 Announce Type: new Abstract: The estimation of odometry in legged robots depends on the assumption that the velocity of the foot with respect to the world remains zero during the st

researcharxiv-cs-ro

Learning from Demonstrations over Riemannian Manifolds using Neural ODEs: An Extended Abstract

arXiv:2606.05422v1 Announce Type: new Abstract: Learning from demonstratins (LfD) is usually performed over Euclidean spaces, while the robot state, e.g. orientation, naturally evolves over curved spa

researcharxiv-cs-ro

Applications

Learning Geometric Representations from Videos for Spatial Intelligent Multimodal Large Language Models

arXiv:2606.05833v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) excel at 2D semantic understanding but lack intrinsic 3D awareness, resulting in representations that fail to m

applicationsarxiv-cs-cv

Learning of Robot Safety Policies via Adversarial Synthetic Scenarios

arXiv:2606.05952v1 Announce Type: new Abstract: In this work, we propose an agentic gamification framework for hazard-informed learning of robot safety policies through synthetic scenarios. We model s

safetyarxiv-cs-ro

Applications

Learning Predictive Visuomotor Coordination

arXiv:2503.23300v2 Announce Type: replace Abstract: Understanding and predicting human visuomotor coordination is crucial for applications in robotics, human-computer interaction, and assistive techno

applicationsarxiv-cs-cv

Tutorials

Learning Self-Correction in Vision-Language Models via Rollout Augmentation

arXiv:2602.08503v2 Announce Type: replace-cross Abstract: Self-correction is essential for solving complex reasoning problems in vision-language models (VLMs). However, existing reinforcement learning

tutorialsarxiv-cs-cl

Learning to Route LLMs from Implicit Cost-Performance Preferences via Meta-Learning

arXiv:2606.06178v1 Announce Type: cross Abstract: Large language models (LLMs) present a trade-off between performance and cost, where more powerful models incur greater expense. LLM routing aims to m

Learning Visual Spatial Planning from Symbolic State via Modality-Gap-Aware Self-Distillation

arXiv:2606.06076v1 Announce Type: cross Abstract: While vision-language models excel at general multimodal understanding, they still struggle with visual spatial planning. We attribute this to a perce

safetyarxiv-cs-cv

Learning What to Forget: Improving LLM Unlearning via Learned Token-Level Importance

arXiv:2606.06320v1 Announce Type: cross Abstract: Machine unlearning aims to remove targeted knowledge from a trained model while preserving its general capabilities. For autoregressive language model

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

Let It Be Simple: One-Step Action Generation for Vision-Language-Action Models

arXiv:2606.05737v1 Announce Type: new Abstract: Diffusion-based vision-language-action (VLA) models often inherit the image-generation view: actions are generated by iterative denoising. We argue that

safetyarxiv-cs-cv

Let’s go 🚀

Let’s go 🚀 Building AI that Builds AI: Introducing the Sakana AI RSI Lab 🚀 https://sakana.ai/rsi-lab Today, we are announcing the Sakana AI Recursive Self-Improvement (RSI) Lab: a dedicated research g

researchdavid-ha--x

Leveraging Large Language Models for Generating Research Topic Ontologies: A Multi-Disciplinary Study

arXiv:2508.20693v2 Announce Type: replace-cross Abstract: Ontologies and taxonomies of research fields are critical for managing and organising scientific knowledge, as they facilitate efficient class

LiAuto-GeoX: Efficient Grounded Driving Transformer

arXiv:2606.05774v1 Announce Type: new Abstract: Dense 3D reconstruction has demonstrated immense potential for spatial understanding, yet its viability as a real-time, onboard representation for auton

Local Ai

Lightricks to split into two companies as it cuts another 75 jobs

local-air-stablediffusion

Lightricks is laying off 75 employees (17% of workforce) as it continues restructuring amid AI disruption, less than six months after a previous round of 85 layoffs. The company is splitting into two

LightVesselNet: An Ultra-Lightweight Sub-100K Parameter Network for Retinal Blood Vessel Segmentation

arXiv:2606.05354v1 Announce Type: new Abstract: Retinal blood vessel segmentation plays a vital role in the early detection of diabetic retinopathy and glaucoma. While recent deep learning models have

Literally the only reason to bailout OpenAI.

Gary Marcus argues for a specific rationale supporting a potential OpenAI bailout, though the exact reasoning is not detailed in the available information. Given Marcus's background as an AI researche

safetygary-marcus--x

LLM-Conditioned Synthesis of Pathological Gaits via Structured Gait-Language Representations

arXiv:2606.06048v1 Announce Type: new Abstract: Pathological gait datasets remain scarce due to privacy, recruitment, cost, and movement variability. Our work presents a multimodal LLM-guided framewor

researcharxiv-cs-cv

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

LLM-Guided ANN Index Optimization for Human-Object Interaction Retrieval

arXiv:2606.05489v1 Announce Type: new Abstract: Retrieval systems underpin modern AI applications -- spanning visual search, recommendation engines, and multi-modal question answering. Modern multi-st

LLMs Can Leak Training Data But Do They Want To? A Propensity-Aware Evaluation of Memorization in LLMs

arXiv:2606.06286v1 Announce Type: new Abstract: Large language models can reproduce training data, but existing memorization evaluations mostly measure whether models can be forced to do so, rather th

Applications

LLMs do not teach people to think critically. Colleges do.

applicationsgary-marcus--x

LLMs do not teach people to think critically. Colleges do. Hot take: Universities charge $300,000 for a degree that teaches you skills any LLM can do for free. At some point we need to have an honest

Localizing Prompt Ambiguity in Large Language Models with Probe-Targeted Attribution

arXiv:2606.05486v1 Announce Type: new Abstract: Prompt ambiguity is a common source of failure in large language models, but is difficult to localize because it is a latent property of the prompt, whi

LongSpace: Exploring Long-Horizon Spatial Memory from Perception to Recall in Video

arXiv:2606.05677v1 Announce Type: cross Abstract: Multimodal Large Language Models (MLLMs) have advanced image and video understanding and can increasingly handle longer visual inputs. Long-horizon ta

LoomVideo: Unifying Multimodal Inputs into Video Generation and Editing

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

LoRi: Low-Rank Distillation for Implicit Reasoning

arXiv:2606.05315v1 Announce Type: new Abstract: Implicit chain-of-thought (iCoT) methods aim to internalize reasoning in large language models, but often underperform explicit CoT prompting. We empiri

Tutorials

Love this. Juan Hernandez immigrates to the US, learns how to weld because it pays well, begins contracting at SpaceX, gets full time role +…

tutorialsemad-mostaque--x

Love this. Juan Hernandez immigrates to the US, learns how to weld because it pays well, begins contracting at SpaceX, gets full time role + $10,000 equity. It’s now worth just under a million dollars

Tools

Luca (@agi2asi) built Grid to centralize 12 different Google Drives into one hub. With Replit, he turned a prompt into an AI-powered employe…

Luca (@agi2asi) built Grid to centralize 12 different Google Drives into one hub. With Replit, he turned a prompt into an AI-powered employee hub in just 10 minutes. Now, team members can ask about an

toolsreplit--x

Tools

Luca built Grid to centralize 12 different Google Drives into one hub. With Replit, he turned a prompt into an AI-powered employee hub in ju…

Luca built Grid to centralize 12 different Google Drives into one hub. With Replit, he turned a prompt into an AI-powered employee hub in just 10 minutes. Now, team members can ask about anything from

toolsreplit--x

Many Circuits, One Mechanism: Input Variation and Evaluation Granularity in Circuit Discovery

arXiv:2606.06267v1 Announce Type: new Abstract: Circuit discovery methods identify subgraphs that explain specific model behaviors, and structural differences between discovered circuits are commonly

MARDoc: A Memory-Aware Refinement Agent Framework for Multimodal Long Document QA

arXiv:2606.05749v1 Announce Type: new Abstract: Iterative retrieval-reasoning agents have recently shown promise for multimodal long-document question answering. However, most existing systems maintai

agentsarxiv-cs-cl

Industry

Marvell and Flex, a contract manufacturer for electronics, will join the S&P 500; MRVL jumps 6%+ after hours after closing down 16.74% amid a broader sell-off (Kif Leswing/CNBC)

Kif Leswing / CNBC: Marvell and Flex, a contract manufacturer for electronics, will join the S&P 500; MRVL jumps 6%+ after hours after closing down 16.74% amid a broader sell-off — - Marvell Technolog

industrytechmeme

MASF: A Multi-Model Adaptive Selection Framework for Abstractive Text summarization

arXiv:2606.05494v1 Announce Type: new Abstract: Automatic text summarization has become increasingly important due to the rapid growth of digital textual information. This paper presents a Multi-Model

Massive output uptick due to agentic AI. Complete flat adoption.

Jeremy Howard discusses a significant increase in output capabilities driven by agentic AI systems, while noting that adoption rates remain completely flat across the board. The observation suggests a

agentsjeremy-howard--x

MAviS: A Multimodal Conversational Assistant For Avian Species

arXiv:2603.07294v2 Announce Type: replace Abstract: Fine-grained understanding and species-specific multimodal question answering are vital for advancing biodiversity conservation and ecological monit

MCBench: A Multicontext Safety Assessment Benchmark for Omni Large Language Models

arXiv:2606.05177v1 Announce Type: new Abstract: Existing multimodal safety benchmarks focus solely on visual inputs and cannot assess Omni Large Language Models (LLMs) that process vision, audio, and

MDP-GRPO: Stabilized Group Relative Policy Optimization for Multi-Constraint Instruction Following

arXiv:2606.06058v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards is ideal for multi-constraint instruction following, yet standard group-relative policy optimization (G

Measuring the sensitivity of LLM-based structured extraction to prompt, model, and schema choices in clinical discharge summaries

arXiv:2606.05970v1 Announce Type: new Abstract: Large language models are increasingly used for structured extraction from clinical free-text notes, but the sensitivity of their output to upstream con

Local Ai

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

MemoryCard: Topic-Aware Multi-Modal Clue Compression for Long-Video Question Answering

arXiv:2606.05917v1 Announce Type: cross Abstract: Long-video question answering remains challenging for Vision-Language Models (VLMs), as answer-relevant evidence is often sparse, transient, and tempo

Merging model-based control with multi-agent reinforcement learning for multi-agent cooperative teaming strategies

arXiv:2606.06011v1 Announce Type: new Abstract: In this work, we propose a framework that combines multi-agent reinforcement learning (MARL) with model-based control to achieve safe, dynamically feasi

agentsarxiv-cs-ro

Meridian: Metric-Semantic Primitive Matching for Cross-View Geo-Localization Beyond Urban Environments

arXiv:2606.06312v1 Announce Type: new Abstract: Successful robot automation requires accurate global localization to support repeatability, task planning, goal specification, and safe operation. Howev

agentsarxiv-cs-ro

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

MLEvolve: A Self-Evolving Framework for Automated Machine Learning Algorithm Discovery

arXiv:2606.06473v1 Announce Type: cross Abstract: Large language model (LLM) agents are increasingly applied to long-horizon tasks such as scientific discovery and machine learning engineering (MLE),

agentsarxiv-cs-cl