All entries

Cognition AI announced a security vulnerability remediation program for Devin, their AI software engineer product, detailing procedures for reporting and addressing security issues. The program establ

agentscognition-ai--x

Local Ai

Readable but Not Controllable: Neuron-Level Evidence for Medical LLM Hallucination

DGX agent

arXiv:2607.00158v1 Announce Type: new Abstract: Hallucination remains one of the central obstacles to deploying medical LLMs. Yet, even when hallucination can be detected, it is still unclear whether

local-aiarxiv-cs-cl

2 Jul 2026

Reading Order Inference for Complex Document Layouts

arXiv:2607.01018v1 Announce Type: cross Abstract: Reading order inference remains a critical bottleneck in the digitization of complex historical manuscripts, where pages contain multiple spatially in

Real-Time Hard Negative Sampling via LLM-based Clustering for Large-Scale Two-Tower Retrieval

arXiv:2607.00448v1 Announce Type: cross Abstract: The two-tower model has been widely used for large-scale recommendation systems, particularly in the retrieval stage. Industry standards for training

safetyarxiv-cs-ai

Reasoning Up the Instruction Ladder for Controllable Language Models

arXiv:2511.04694v5 Announce Type: replace-cross Abstract: As large language model (LLM) based systems take on high-stakes roles in real-world decision-making, they must reconcile competing instruction

safetyarxiv-cs-ai

Recovering Input Text from Hidden States: Study of Gradient-Based Inversion of Decoder-Only Language Models

model-releasesarxiv-cs-ai

arXiv:2607.00852v1 Announce Type: cross Abstract: This work studies the hidden-state inversion problem: recovering the original input token sequence of a decoder-only language model from its last-laye

Recursively Enumerably Representable Classes and Computable Versions of the Fundamental Theorem of Statistical Learning

arXiv:2511.02644v2 Announce Type: replace Abstract: We study computable probably approximately correct (CPAC) learning, where learners are required to be computable functions. It had been previously o

Relation-Centric Open-Vocabulary 3D Gaussian Segmentation

arXiv:2607.01140v1 Announce Type: new Abstract: Open-vocabulary 3D Gaussian segmentation is challenging because it requires language understanding for diverse queries and accurate separation of Gaussi

researcharxiv-cs-cv

Representation as a Bottleneck for Mechanistic Interpretability: The Manifestation Unit Protocol

arXiv:2607.00089v1 Announce Type: new Abstract: Mechanistic interpretability has produced a rich inventory of component-level analyses that characterise what neural-network components encode and how t

Residual Context Diffusion Language Models

researchapple-ml-research

Diffusion Large Language Models (dLLMs) have emerged as a promising alternative to purely autoregressive language models because they can decode multiple tokens in parallel. However, state-of-the-art

Tutorials

Restore3D: Breathing Life into Broken Objects with Shape and Texture Restoration

arXiv:2607.00522v1 Announce Type: new Abstract: Restoring incomplete or damaged 3D objects is crucial for cultural heritage preservation, occluded object reconstruction, and artistic design. Existing

tutorialsarxiv-cs-cv

RetailSMV: Exocentric vs. Egocentric Adaptation of Foundation Video World Models in Retail

model-releasesarxiv-cs-ai

arXiv:2607.00310v1 Announce Type: cross Abstract: Foundation video diffusion models are increasingly viewed as world simulators for embodied agents, yet their pretraining on internet-scale generic vid

Agents

Rethinking Multi-Label Image Classification With Deep Learning: Taxonomy, Challenge, and Outlook

arXiv:2607.00839v1 Announce Type: new Abstract: Multi-label image classification (MLIC), a fundamental task in computer vision, focuses on identifying multiple objects or concepts within an image, und

agentsarxiv-cs-cv

Rethinking Robust Adversarial Concept Erasure in Diffusion Models

arXiv:2510.27285v4 Announce Type: replace Abstract: Concept erasure methods aim to remove specific unsafe target concepts in diffusion models while preserving image generation utility. To address the

researcharxiv-cs-cv

Rethinking Visual Privacy: A Compositional Privacy Risk Framework for Severity Assessment with VLMs

arXiv:2603.21573v2 Announce Type: replace Abstract: Existing visual privacy benchmarks largely treat privacy as a binary property, labeling images as private or non-private based on visible sensitive

researcharxiv-cs-cv

Retrieved Images as Visual Thought: Training-Free Multimodal In-Context Learning for the Open-vs-Closed Gap

arXiv:2607.00606v1 Announce Type: new Abstract: Recent work on Thinking with Images makes vision a dynamic part of reasoning, but does so through generation: the model invokes external tools, synthesi

Revisiting Autoregressive Models for Generative Image Classification

arXiv:2603.19122v2 Announce Type: replace Abstract: Class-conditional generative models have emerged as accurate and robust classifiers, with diffusion models demonstrating clear advantages over other

safetyarxiv-cs-cv

Reward function compression facilitates goal-dependent reinforcement learning

arXiv:2509.06810v3 Announce Type: replace-cross Abstract: Humans can uniquely assign value to novel, abstract outcomes to support reinforcement learning. However, this flexibility is cognitively costl

Right in the Right Way: LM Training with Verifiable Rewards and Human Demonstrations

model-releasesarxiv-cs-ai

arXiv:2607.01181v1 Announce Type: cross Abstract: RL with verifiable rewards (RLVR) has emerged as a powerful paradigm for training LMs on tasks with well-defined success metrics, such as code generat

RoadBench: Benchmarking MLLMs on Fine-Grained Spatial Understanding and Reasoning under Urban Road Scenarios

arXiv:2511.18011v2 Announce Type: replace Abstract: Multimodal large language models (MLLMs) have demonstrated powerful capabilities in general spatial understanding and reasoning. However, their fine

Robots Ask the Way: Communication-Enabled Social Navigation

arXiv:2607.01044v1 Announce Type: new Abstract: Assistive autonomous robots operating in multi-agent environments require efficient strategies to locate specific individuals among multiple residents.

safetyarxiv-cs-ro

RoboWorld: Fast and Reliable Neural Simulators for Generalist Robot Policy Evaluation

arXiv:2607.01060v1 Announce Type: new Abstract: Video world models are emerging as a scalable alternative for evaluating generalist robot policies, bypassing the physical constraints and engineering b

safetyarxiv-cs-ro

Robust 3D Alignment of Generative Reconstructions via Partial Monocular Observations

arXiv:2607.00498v1 Announce Type: new Abstract: Aligning generative 3D reconstructions with partial monocular observations is a critical but under-explored challenge in computer vision. This task is i

Robust Operational Space Control with Conformal Disturbance Bounds for Safe Redundant Manipulation

arXiv:2607.00424v1 Announce Type: new Abstract: Redundant robotic manipulators operating in constrained and human-interactive environments require accurate task-space tracking together with rigorous s

safetyarxiv-cs-ro

Root Cause Analysis of Outliers in Unknown Cyclic Graphs

arXiv:2510.06995v2 Announce Type: replace-cross Abstract: We study the propagation of outliers in cyclic causal graphs with linear structural equations, tracing them back to one or several 'root cause

Hardware

ROSA: A Robotics Foundation Model Serving System for Robot Factories

arXiv:2607.01088v1 Announce Type: new Abstract: Robotics foundation models (RFMs) are making general-purpose robots increasingly practical for factory deployments. While RFM serving systems are centra

hardwarearxiv-cs-ro

Rosetta: Composable Native Multimodal Pretraining

arXiv:2607.00293v1 Announce Type: cross Abstract: Achieving true artificial general intelligence requires foundation models capable of integrating new modalities without forgetting prior knowledge. Ho

researcharxiv-cs-cl

Tools

Routing rules now available on AI Gateway

Vercel has introduced routing rules functionality for AI Gateway, enabling users to define custom logic for directing requests to different AI models or endpoints. This feature allows developers to im

toolsvercel-blog

Safe Alone, Unsafe Together: Safeguarding Against Implicit Toxicity When Benign Images Combine

arXiv:2607.00576v1 Announce Type: new Abstract: Multi-image content has become an increasingly prevalent form of visual communication in social media, giving rise to a new safety issue, multi-image im

safetyarxiv-cs-cl

Salt: Self-Consistent Distribution Matching with Cache-Aware Training for Fast Video Generation

arXiv:2604.03118v2 Announce Type: replace Abstract: Distilling video generation models to extremely low inference budgets (e.g., 2--4 NFEs) is crucial for real-time deployment, yet remains challenging

safetyarxiv-cs-cv

Sam Altman calls for US-led international forum to set global AI standards

Sam Altman is calling for a US-led international forum to set global safety standards for artificial intelligence, arguing that no single country should be left to dominate the technology. In an op-ed

safetysiliconangle

Industry

Sam Altman proposes a 'US-led international forum' to establish AI standards, provide analysis of capabilities and risks, and make AI available to allies (Sam Altman/Financial Times)

Sam Altman / Financial Times: Sam Altman proposes a “US-led international forum” to establish AI standards, provide analysis of capabilities and risks, and make AI available to allies — The labs devel

industrytechmeme

Sample Complexities of Estimating Gumbel--Max Watermark Proportions with and without Reduction to Pivotal Statistics

arXiv:2607.00224v1 Announce Type: cross Abstract: Watermarking promises a statistical trace of large language model (LLM) use, but real documents, after editing or paraphrasing, rarely arrive as purel

SAOT: Self-Supervised Continual Graph Learning with Structure-Aware Optimal Transport

model-releasesarxiv-cs-lg

arXiv:2607.00377v1 Announce Type: new Abstract: Self-supervised Continual Graph Learning (CGL) aims to successively learn from a graph sequence with different tasks without label supervision - a parad

Scaling Up Thermodynamic AI Models

arXiv:2607.00170v1 Announce Type: cross Abstract: Thermodynamic computing devices based on the Ising model show great promise for low-power AI inference and edge computing, but scalable methods for tr

scDataset: Scalable Data Loading for Deep Learning on Large-Scale Single-Cell Omics

arXiv:2506.01883v3 Announce Type: replace-cross Abstract: Training deep learning models on single-cell datasets with hundreds of millions of cells requires loading data from disk, as these datasets ex

Applications

SchemaRAG: Dynamic Large Schema Reduction for LLM-driven Structured Information Extraction

arXiv:2607.00008v1 Announce Type: cross Abstract: Extracting structured data from unstructured text using large language models (LLMs) becomes challenging when target schemas are large and complex. In

applicationsarxiv-cs-ai

Applications

SD-RouteFusion: Ego-Trajectory Prediction with SD-Map Route Conditioning

arXiv:2607.01139v1 Announce Type: new Abstract: This paper presents SD-RouteFusion, a deployable end-to-end ego-trajectory prediction method that fuses a front-facing camera, vehicle kinematics, and a

applicationsarxiv-cs-cv

Seahorse: A Unified Benchmarking Framework for Spatiotemporal Event Modeling

model-releasesarxiv-cs-lg

arXiv:2607.01022v1 Announce Type: new Abstract: Spatiotemporal point processes (STPPs) model event data in continuous time and space, with applications in mobility, epidemiology, and public safety. Re

Search-Based Spatiotemporal and Multi-Robot Motion Planning on Graphs of Space-Time Convex Sets

arXiv:2607.00444v1 Announce Type: cross Abstract: Spatiotemporal motion planning, especially in multi-robot settings, requires robots to reason about collision-free regions that change over time, whic

Tools

See how our latest funding round accelerates the shift to open-source AI: https://www.together.ai/blog/announcing-our-series-c

Together AI announced a Series C funding round aimed at accelerating the adoption and development of open-source AI models and infrastructure. The funding enables the company to expand its platform fo

toolstogether-ai--x

Applications

Seed2.0 Model Card: Towards Intelligence Frontier for Real-World Complexity

arXiv:2607.00248v1 Announce Type: new Abstract: We present Seed2.0, a model series that takes a meaningful step toward solving complex, real-world tasks. Our approach begins with identifying users' ge

applicationsarxiv-cs-ai

SEFORA: Student Essays with Feedback Corpus and LLM Feedback Evaluation Framework

arXiv:2607.00274v1 Announce Type: cross Abstract: Effective writing feedback is among the strongest drivers of student learning, yet producing it at scale is labor-intensive. LLMs offer a natural path

SegFly: A Dataset and 2D-3D-2D Paradigm for Aerial RGB-Thermal Semantic Segmentation at Scale

arXiv:2603.17920v2 Announce Type: replace Abstract: Semantic segmentation for uncrewed aerial vehicles (UAVs) is fundamental for aerial scene understanding, yet existing RGB and RGB-T datasets remain

Local Ai

Segmenting, Fast and Slow: Real-Time Open-Vocabulary Video Instance Segmentation with Dual-Path Processing

arXiv:2607.00124v1 Announce Type: new Abstract: Object-centric models inspired by DETR have become the dominant paradigm for open-vocabulary video instance segmentation (OV-VIS). While recent efforts

local-aiarxiv-cs-cv

Selective Expert Guidance for Effective and Diverse Exploration in Reinforcement Learning of LLMs

arXiv:2510.04140v2 Announce Type: replace Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has become a widely adopted technique for enhancing the reasoning ability of Large Language Mo

safetyarxiv-cs-ai

Selective Test-Time Debiasing for CLIP via Reward Gating

arXiv:2607.00423v1 Announce Type: new Abstract: Vision language models (VLMs) demonstrate strong zero-shot performance, but often perpetuate social stereotypes in person-centric queries, yielding skew

safetyarxiv-cs-cl

Tutorials

Self-conditioned Flow Map Language Models via Fixed-point Flows

arXiv:2607.00714v1 Announce Type: cross Abstract: Self-conditioning is a core technique that enhances continuous flow-based language models, where the model learns to denoise generated text by conditi

tutorialsarxiv-cs-ai