Search: “models”

arXiv:2507.21183v5 Announce Type: replace-cross Abstract: As the era of large language models (LLMs) unfolds, Preference Optimization (PO) methods have become a central approach to aligning LLMs with

safetyarxiv-cs-ai

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Mask2Cause: Causal Discovery via Adjacency Constrained Causal Attention

DGX agent

arXiv:2605.07280v1 Announce Type: cross Abstract: Leveraging deep learning for causal discovery in time series remains challenging because existing neural methods predominantly rely on component-wise

model-releasesarxiv-cs-ai

11 May 2026

Mathematical Reasoning via Intervention-Based Time-Series Causal Discovery Using LLMs as Concept Mastery Simulators

arXiv:2605.07600v1 Announce Type: cross Abstract: Recent methods for improving LLM mathematical reasoning, whether through MCTS-based test-time search or causal graph-guided knowledge injection, canno

McNdroid: A Longitudinal Multimodal Benchmark for Robust Drift Detection in Android Malware

model-releasesarxiv-cs-lg

arXiv:2605.06894v1 Announce Type: cross Abstract: Machine learning (ML) in real-world systems must contend with concept drift, adversarial actors, and a spectrum of potential features with varying cos

Meet the latest Database Center, now with Gemini-powered fleet intelligence

model-releasesgoogle-cloud-ai

Managing a modern database fleet is both a scale and cognitive problem. As database estates grow, the effort required to monitor, troubleshoot, and optimize them often outpaces teams’ capacity, who fi

MIPIAD: Multilingual Indirect Prompt Injection Attack Defense with Qwen -- TF-IDF Hybrid and Meta-Ensemble Learning

arXiv:2605.07269v1 Announce Type: new Abstract: Indirect prompt injection remains a persistent weakness in retrieval-augmented and tool-using LLM systems, and the problem becomes harder to characteris

Multi-Objective Constraint Inference using Inverse reinforcement learning

arXiv:2605.06951v1 Announce Type: new Abstract: Constraint inference is widely considered essential to align reinforcement learning agents with safety boundaries and operational guidelines by observin

Neural Operators as Efficient Function Interpolators

arXiv:2605.07792v1 Announce Type: cross Abstract: Neural operators (NOs) are designed to learn maps between infinite-dimensional function spaces. We propose a novel reframing of their use. By introduc

New TIL: I figured out how to use my LLM CLI tool in a shebang line, which means you can write executable scripts in English, or hook up mor…

model-releasessimon-willison--x

Simon Willison discovered how to use an LLM command-line interface tool in Unix shebang lines, enabling the creation of executable scripts written in English or natural language. This technique allows

NPMixer: Hierarchical Neighboring Patch Mixing for Time Series Forecasting

model-releasesarxiv-cs-lg

arXiv:2605.07476v1 Announce Type: new Abstract: Multivariate time series forecasting remains a challenge due to the complexity of local temporal dynamics and global dependencies across multiple variab

On the Robustness of Distribution Support under Diffusion Guidance

arXiv:2605.07220v1 Announce Type: new Abstract: Diffusion guidance is a powerful technique that enables controllable and high-fidelity sample generation with diffusion models. At a high level, it modi

researcharxiv-cs-lg

Agents

On Time, Within Budget: Constraint-Driven Online Resource Allocation for Agentic Workflows

arXiv:2605.06110v2 Announce Type: replace Abstract: Agentic systems increasingly solve complex user requests by executing orchestrated workflows, where subtasks are assigned to specialized models or t

agentsarxiv-cs-ai

OpenAI Campus Network: Student club interest form

The OpenAI Campus Network is a program that facilitates student engagement with OpenAI's technology and research on college campuses. This interest form allows students to express interest in starting

model-releasesopenai

OpenAI just released its answer to Claude Mythos

model-releasesthe-verge-ai

OpenAI is launching Daybreak, an AI initiative focused on detecting and patching vulnerabilities before attackers find them. Daybreak uses the Codex Security AI agent that launched in March to create

OrScale: Orthogonalised Optimization with Layer-Wise Trust-Ratio Scaling

arXiv:2605.07815v1 Announce Type: cross Abstract: Muon improves neural-network training by orthogonalizing matrix-valued updates, but it leaves each layer's update magnitude controlled mostly by a glo

Tutorials

Path Integration and Object-Location Binding Emerge in an Action-Conditioned Predictive Sequence Network

arXiv:2602.03490v2 Announce Type: replace Abstract: Adaptive cognition requires structured internal models of objects and their relations. Predictive neural networks are often proposed to learn such w

tutorialsarxiv-cs-lg

PRIMED: Adaptive Modality Suppression for Referring Audio-Visual Segmentation via Biased Competition

model-releasesarxiv-cs-cv

arXiv:2605.07154v1 Announce Type: new Abstract: Referring Audio-Visual Segmentation (Ref-AVS) seeks to localize and segment target objects in video frames based on visual, auditory, and textual referr

Tutorials

Prototype Guided Post-pretraining for Single-Cell Representation Learning

arXiv:2605.07938v1 Announce Type: new Abstract: Single-cell representation learning (SCRL) from gene expression data offers a way to uncover the complex regulatory logic underlying cellular function.

tutorialsarxiv-cs-lg

ProtSent: Protein Sentence Transformers

arXiv:2605.06830v1 Announce Type: cross Abstract: Protein language models (pLMs) produce per-residue representations that capture evolutionary and structural information, yet their mean-pooled sequenc

researcharxiv-cs-cl

RCoT-Seg: Reinforced Chain-of-Thought for Video Reasoning and Segmentation

model-releasesarxiv-cs-cv

arXiv:2605.07334v1 Announce Type: new Abstract: Video Reasoning Segmentation (VRS) aims to segment target objects in videos based on implicit instructions that convey human intent and temporal logic.

Safety

Reason to Play: Behavioral and Brain Alignment Between Frontier LRMs and Human Game Learners

arXiv:2605.08019v1 Announce Type: new Abstract: Humans rapidly learn abstract knowledge when encountering novel environments and flexibly deploy this knowledge to guide efficient and intelligent actio

safetyarxiv-cs-ai

Region4Web: Rethinking Observation Space Granularity for Web Agents

arXiv:2605.07134v1 Announce Type: cross Abstract: Web agents perceive web pages through an observation space, yet its granularity has remained an underexamined design choice. Existing work treats obse

RL-RIG: A Generative Spatial Reasoner via Intrinsic Reflection

arXiv:2602.19974v2 Announce Type: replace Abstract: Recent advancements in image generation have achieved impressive results in producing high-quality images. However, existing image generation models

researcharxiv-cs-cv

SatSurfGS: Generalizable 2D Gaussian Splatting for Sparse-View Satellite Surface Reconstruction

model-releasesarxiv-cs-cv

arXiv:2605.07181v1 Announce Type: new Abstract: Sparse-view satellite image surface reconstruction remains highly challenging, fundamentally because the reliability of multi-view matching under satell

ScrapeGraphAI-100k: Dataset for Schema-Constrained LLM Generation

arXiv:2602.15189v2 Announce Type: replace-cross Abstract: Producing output that conforms to a specified JSON schema underlies tool use, structured extraction, and knowledge base construction in modern

Local Ai

Search-based Robustness Testing of Laptop Refurbishing Robotic Software

arXiv:2605.07530v1 Announce Type: new Abstract: The Danish Technological Institute (DTI) focuses on transferring advanced technologies (including robots) to the industry and the public sector. One key

local-aiarxiv-cs-ro

Semantic Integrity Matters: Benchmarking and Preserving High-Density Reasoning in KV Cache Compression

arXiv:2502.01941v3 Announce Type: replace-cross Abstract: While Key-Value (KV) cache compression is essential for efficient LLM inference, current evaluations disproportionately focus on sparse retrie

SlopCodeBench: Benchmarking How Coding Agents Degrade Over Long-Horizon Iterative Tasks

arXiv:2603.24755v2 Announce Type: replace-cross Abstract: Software development is iterative, yet agentic coding benchmarks hide design issues through their single-shot setup. Recent iterative benchmar

SREGym: A Live Benchmark for AI SRE Agents with High-Fidelity Failure Scenarios

arXiv:2605.07161v1 Announce Type: new Abstract: AI agents are increasingly used to diagnose and mitigate failures in production systems, known as agentic Site Reliability Engineering (SRE). Current SR

SteelDefectX: A Multi-Form Vision-Language Dataset and Benchmark for Steel Surface Defect Analysis

arXiv:2603.21824v2 Announce Type: replace-cross Abstract: Steel surface defect analysis is critical for industrial quality control, yet existing benchmarks rely primarily on label-only annotations, li

Tutorials

Structured Coupling for Flow Matching

arXiv:2605.07676v1 Announce Type: new Abstract: Standard flow matching scales well but typically relies on an unstructured source distribution, limiting its ability to learn interpretable latent struc

tutorialsarxiv-cs-lg

The EDelta-MHC-Geo Transformer: Adaptive Geodesic Operations with Guaranteed Orthogonality

arXiv:2605.06729v1 Announce Type: cross Abstract: We present the EDelta-MHC-Geo Transformer, a novel architecture that unifies Manifold-Constrained Hyper-Connections (mHC), Deep Delta Learning (DDL),

The Memory Curse: How Expanded Recall Erodes Cooperative Intent in LLM Agents

arXiv:2605.08060v1 Announce Type: cross Abstract: Context window expansion is often treated as a straightforward capability upgrade for LLMs, but we find it systematically fails in multi-agent social

Topic Is Not Agenda: A Citation-Community Audit of Text Embeddings

arXiv:2605.07158v1 Announce Type: cross Abstract: Vector search and retrieval-augmented generation (RAG) rest on the assumption that cosine similarity between text embeddings reflects conceptual relat

Agents

TraceFix: Repairing Agent Coordination Protocols with TLA+ Counterexamples

arXiv:2605.07935v1 Announce Type: new Abstract: We present TraceFix, a verification-first pipeline for Large Language Model (LLM) multi-agent coordination. An agent synthesizes a protocol topology as

agentsarxiv-cs-ai

Training-Induced Escape from Token Clustering in a Mean-Field Formulation of Transformers

model-releasesarxiv-cs-lg

arXiv:2605.07772v1 Announce Type: new Abstract: Transformers perform inference by iteratively transforming token representations across layers. This layerwise computation has been studied empirically,

Using LLM in the shebang line of a script

model-releasessimon-willison

TIL: Using LLM in the shebang line of a script Kim_Bruning on Hacker News: But seriously, you can put a shebang on an english text file now (if you're sufficiently brave) [...] This inspired me to loo

Utility-Preserving De-Identification for Math Tutoring: Investigating Numeric Ambiguity in the MathEd-PII Benchmark Dataset

arXiv:2602.16571v2 Announce Type: replace Abstract: Large-scale sharing of dialogue data is key to advancing the science of teaching and learning, yet rigorous de-identification remains a major barrie

Visual Text Compression as Measure Transport

arXiv:2605.06708v1 Announce Type: cross Abstract: Visual text compression (VTC) promises efficient long-context processing by rendering text into an image and re-encoding it with a vision-language mod

📣We're calling for ambassadors! Whether you're a developer with great technical taste or a local community leader who loves bringing people…

📣We're calling for ambassadors! Whether you're a developer with great technical taste or a local community leader who loves bringing people together, we'd love to have you join us. Visit the website b

model-releasesqwen--x

When Losses Align: Gradient-Based Composite Loss Weighting for Efficient Pretraining

arXiv:2605.07756v1 Announce Type: cross Abstract: Modern deep models are often pretrained on large-scale data with missing labels using composite objectives, where the relative weights of multiple los

researcharxiv-cs-ai

When Routine Chats Turn Toxic: Unintended Long-Term State Poisoning in Personalized Agents

arXiv:2605.06731v1 Announce Type: cross Abstract: Personalized LLM agents maintain persistent cross-session state to support long-horizon collaboration. Yet, this persistence introduces a subtle but c

Zero-Shot Satellite Image Retrieval through Joint Embeddings: Application to Crisis Response

arXiv:2605.05405v2 Announce Type: replace Abstract: Semantic search of Earth observation archives remains challenging. Visual foundation models such as CLAY produce rich embeddings of satellite imager

researcharxiv-cs-cv

@roydanroy What he talks about couldn't have happened before GPT-5.5

model-releasessam-altman--x

Sam Altman comments on a statement by @roydanroy, suggesting that the technological capabilities or developments being discussed would not have been possible prior to GPT-5.5. The post implies that GP

10 May 2026

Downloading now... 1M token context window with supposedly usable coding agent capability all on a 128GB Macbook Pro is 🤯

model-releasesclem-delangue--x

Downloading now... 1M token context window with supposedly usable coding agent capability all on a 128GB Macbook Pro is 🤯 🚨 OPEN SOURCE AI IS LITERALLY UNSTOPPABLE 🚨 The legendary founder of Redis (An

Hot take on METR’s new graph that so many people are flipping about today. • Claude Code is a real advance; Mythos probably builds on some o…

model-releasesgary-marcus--x

Hot take on METR’s new graph that so many people are flipping about today. • Claude Code is a real advance; Mythos probably builds on some of what is learned there. But… • If you read the graph carefu

This is confused, but popular. Popular because it tells a bunch of people what they want to hear. Confused for a couple reasons: first, Myth…

model-releasesgary-marcus--x

This is confused, but popular. Popular because it tells a bunch of people what they want to hear. Confused for a couple reasons: first, Mythos probably isn’t a pure LLM. (Claude Code isn’t, and it pro

Hardware

Yann LeCun closed $1.03B for AMI Labs on March 10. Three days later, this paper dropped from his NYU collaborators. 15M parameters. Single G…

Yann LeCun closed 1.03B for AMI Labs on March 10. Three days later, this paper dropped from his NYU collaborators. 15M parameters. Single GPU. A few hours of training. LeWorldModel is the first JEPA t

hardwareyann-lecun--x