Search: “models”

How Much Dense Attention is Necessary? Oracle-Guided Sparse Prefill for Full/GQA Layers in Hybrid Long-Context Models

arXiv:2606.07703v1 Announce Type: cross Abstract: Long-context prefill remains expensive because full/GQA layers still score the historical sequence, even in hybrid models with local, sparse, linear,

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

How Well Do Latent World Models Understand Partially Observable Safety Constraints?

DGX agent

arXiv:2510.06492v2 Announce Type: replace Abstract: Latent world models are a promising approach for learning state representations and dynamics directly from high-dimensional observations, enabling r

safetyarxiv-cs-ro

9 Jun 2026

IDEQ -- Improving Diffusion Models for the Traveling Salesman Problem (TSP) by Leveraging the Structure of the Solution Space

arXiv:2412.13858v2 Announce Type: replace Abstract: We investigate diffusion models to solve the Traveling Salesman Problem. Building on the recent DIFUSCO and T2TCO approaches, we propose IDEQ. IDEQ

IMUG-Bench: Benchmarking Unified Multimodal Models on Interleaved Understanding and Generation

arXiv:2606.09169v1 Announce Type: new Abstract: In recent years, unified multimodal models (UMMs) have emerged to support both understanding and generation within a single framework. Mastering dynamic

LargeMonitor: Monitoring Online Task-Free Continual Learning via Large Pretrained Models

arXiv:2606.09430v1 Announce Type: cross Abstract: Online task-free continual learning (TFCL) requires intelligent agents to sequentially accumulate knowledge from an unbounded, non-stationary data str

Late-Layer Fusion is Enough: Dual-Path Vision Token Routing for Multimodal Large Language Models under Visual Saturation

arXiv:2606.09131v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) commonly inherit the deep, symmetric Transformer backbone designed for unimodal text modeling, and apply the sa

Pretrained, Frozen, Still Leaking: Auditing Cross-Encoder Attribute Transfer in EEG Foundation Models

arXiv:2606.09189v1 Announce Type: cross Abstract: EEG foundation-model releases are usually audited one endpoint at a time: raw-reconstruction, membership inference, identity linkage, or DP-SGD on the

PROBE-Web: An Interactive System for Probing Evaluation Landscapes of Knowledge Graph Completion Models

arXiv:2606.08926v1 Announce Type: new Abstract: Knowledge graph completion (KGC) models are commonly evaluated using rank-based metrics such as MRR and Hits@K, despite different users often requiring

safetyarxiv-cs-lg

ProbeAct: Probe-Guided Training-Free Failure Recovery in Vision-Language-Action Models

model-releasesarxiv-cs-ro

arXiv:2606.09740v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models demonstrate strong perfor-1 mance on language-conditioned robotic manipulation within their training dis-2 tribution

SlideCheck: Guiding Self-Supervised Pretraining of Pathology Foundation Models via Dataset Distributions

arXiv:2606.07590v1 Announce Type: cross Abstract: Pathology foundation models are pretrained on large streams of WSI-derived patches, while supervision during data construction is often slide-level, s

Sycophancy as a Multilingual Alignment Failure: How Safety Degrades Across Languages, Topics, and Models

arXiv:2606.08451v1 Announce Type: cross Abstract: Safety-aligned large language models often exhibit sycophancy, which is the tendency to affirm users' opinions regardless of factual accuracy. Althoug

Targeting World Models to Compromise Robot Learning Pipelines

arXiv:2606.09499v1 Announce Type: cross Abstract: World models have recently seen a rapid growth in both their popularity and capability as more data efficient tools for generating robot training data

Video Understanding by Design: How Datasets Shape Video Models

arXiv:2509.09151v2 Announce Type: replace-cross Abstract: Research in video understanding has advanced rapidly, driven by increasingly diverse datasets and more powerful model architectures. While exi

LARA: Latent Action Representation Alignment for Vision-Language-Action Models

arXiv:2606.07100v1 Announce Type: new Abstract: Visual-language action (VLA) models enable robots to predict actions directly from observations and language instructions, but their performance depends

safetyarxiv-cs-cv

Latent-space Attacks for Refusal Evasion in Language Models

arXiv:2605.21706v2 Announce Type: replace Abstract: Safety-aligned language models are trained to refuse harmful requests, yet refusal behavior can be suppressed by steering their internal representat

Applications

SpectCount: Spectrotemporal Counting via Synthetic Signals Improves Large Audio Language Models

arXiv:2606.06907v1 Announce Type: cross Abstract: Large audio language models (LALMs) extend large language models with an audio encoder and large-scale audio data. However, the scarcity of high-quali

applicationsarxiv-cs-ai

Step-Wise Refusal Dynamics in Autoregressive and Diffusion Language Models

arXiv:2602.02600v3 Announce Type: replace-cross Abstract: Diffusion language models (DLMs) have recently emerged as a competitive alternative to autoregressive (AR) models, offering parallel decoding,

Textual Supervision Enhances Geospatial Representations in Vision-Language Models

arXiv:2606.07172v1 Announce Type: cross Abstract: Geospatial understanding is a critical yet underexplored dimension in the development of machine learning systems for tasks such as image geolocation

The Dual Mechanisms of Spatial Variable Binding in Vision-Language Models

arXiv:2603.22278v2 Announce Type: replace Abstract: Many multimodal tasks, such as image captioning and visual question answering, require vision-language models (VLMs) to bind objects with their prop

researcharxiv-cs-cv

Tutorials

In-Training Defenses against Emergent Misalignment in Language Models

arXiv:2508.06249v3 Announce Type: replace-cross Abstract: Fine-tuning lets practitioners repurpose aligned large language models (LLMs) for new domains, yet recent work reveals emergent misalignment (

tutorialsarxiv-cs-ai

6 Jun 2026

Pattern Selectivity is Not Task-Causal Structure: A Cross-Architecture Mechanistic Study of Composed-Task Circuits in 1B-Class Language Models

arXiv:2606.05378v1 Announce Type: cross Abstract: We test whether a single screen-and-ablate recipe -- identify attention-head circuits by task-pattern selectivity, then verify by causal ablation agai

6 Jun 2026

Applications

Also, a lot depends on Chinese labs continuing to ship open weights models. If they stop, the frontier falls further and further behind to t…

applicationsethan-mollick--x

Also, a lot depends on Chinese labs continuing to ship open weights models. If they stop, the frontier falls further and further behind to those who want to use local/fine-tuned models. I think this i

Channel-Wise Mixed-Precision Quantization for Large Language Models

model-releasesarxiv-cs-cl

arXiv:2410.13056v4 Announce Type: replace Abstract: Large Language Models (LLMs) have demonstrated remarkable success across a wide range of language tasks, but their deployment on edge devices remain

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

model-releasesarxiv-cs-cl

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

Sources say xAI used Claude models for distillation and training, including using personal accounts and the intermediary service Blackbox AI after being cut off (Grace Kay/The Information)

Grace Kay / The Information: Sources say xAI used Claude models for distillation and training, including using personal accounts and the intermediary service Blackbox AI after being cut off — SpaceX's

model-releasestechmeme

The Invisible Hand of Physics: When Video Diffusion Models Know More Than They Show

arXiv:2606.05328v1 Announce Type: cross Abstract: Modern video diffusion models generate increasingly realistic and temporally coherent videos, motivating their use as candidate world simulators. Yet

researcharxiv-cs-cv

VOLD: Reasoning Transfer from LLMs to Vision-Language Models via On-Policy Distillation

arXiv:2510.23497v3 Announce Type: replace Abstract: Training vision-language models (VLMs) for complex reasoning remains a challenging task, i.a. due to the scarcity of high-quality image-text reasoni

safetyarxiv-cs-cv

Automatic Generation of Titles for Research Papers Using Language Models

arXiv:2606.05085v1 Announce Type: cross Abstract: The title of a research paper conveys its primary idea and, occasionally, its conclusions in a clear and concise manner. Choosing an appropriate title

CLAW: Learning Continuous Latent Action World Models via Adversarial Latent Regularization

arXiv:2606.04130v1 Announce Type: new Abstract: We introduce CLAW, a fully end-to-end self-supervised framework for learning a world model jointly with continuous latent action representations directl

researcharxiv-cs-ro

Evaluating Zero-Shot and One-Shot Adaptation of Small Language Models in Leader-Follower Interaction

arXiv:2602.23312v3 Announce Type: replace-cross Abstract: Leader-follower interaction is an important paradigm in human-robot interaction (HRI). Yet, assigning roles in real time remains challenging f

EvoPrompt: Guided Prompt Evolution for Vision-Language Models Adaptation

arXiv:2603.09493v2 Announce Type: replace-cross Abstract: The adaptation of large-scale vision-language models (VLMs) to downstream tasks with limited labeled data remains a significant challenge. Whi

Tutorials

Geometry-Preserving Encoder/Decoder in Latent Generative Models

arXiv:2501.09876v3 Announce Type: replace-cross Abstract: Generative modeling aims to generate new data samples that resemble a given dataset. When using diffusion models for this task, one of the mai

tutorialsarxiv-cs-lg

Introducing two NVIDIA Nemotron models on Together AI: Nemotron 3 Ultra for high-throughput agentic workloads and Nemotron 3.5 ASR for low-l…

model-releasestogether-ai--x

Introducing two NVIDIA Nemotron models on Together AI: Nemotron 3 Ultra for high-throughput agentic workloads and Nemotron 3.5 ASR for low-latency multilingual speech recognition. AI natives can now b

Learning Control-Affine Reduced-Order Models via Autoencoders

arXiv:2606.05045v1 Announce Type: cross Abstract: We present in this paper a framework for the identification of control-affine reduced-order models (ROMs). The proposed method utilizes autoencoders (

researcharxiv-cs-lg

Learning Long Range Spatio-Temporal Representations over Continuous Time Dynamic Graphs with State Space Models

arXiv:2606.04672v1 Announce Type: cross Abstract: Continuous-time dynamic graphs (CTDGs) provide a richer framework to capture fine-grained temporal patterns in evolving relational data. Long-range in

New Benchmarking Shows Limited Generalization Power of TCR Antigenic Epitope Prediction Models

model-releasesarxiv-cs-lg

arXiv:2606.04994v1 Announce Type: new Abstract: Accurate computational prediction of T cell receptor (TCR) antigen specificity would transform the study of T cell biology and enable scalable immune en

Pulled the trigger today and switched 100% of Lindy traffic to DeepSeek v4, churning from Anthropic models. Saves us millions of $ and we're…

model-releasesclem-delangue--x

Pulled the trigger today and switched 100% of Lindy traffic to DeepSeek v4, churning from Anthropic models. Saves us millions of $ and we're actually seeing an *increase* in performance on many core u

That's a badass title, and it's true! Every day, it gets harder and harder to create tests that AI models can't beat. Reality is humanity's …

That's a badass title, and it's true! Every day, it gets harder and harder to create tests that AI models can't beat. Reality is humanity's real last exam. Andon Labs' Real-World AI Evals: Claude call

model-releasesswyx--x

Value Entanglement: Conflation Between Different Kinds of Good In (Some) Large Language Models

arXiv:2602.19101v2 Announce Type: replace-cross Abstract: Value alignment of Large Language Models (LLMs) requires us to empirically measure these models' actual, acquired representation of value. Amo

Hardware

Announcing Foundry Managed Compute: Run open models in Microsoft Foundry

hardwaremicrosoft-foundry

Microsoft Foundry Managed Compute is a new GPU platform-as-a-service for hosting open-source and custom AI models behind the same endpoint, SDKs, and bill as frontier models. The post Announcing Found

Applications

AugMask: Training Diffusion Models on Incomplete Tabular Data via Stochastic Augmentation and Masking

arXiv:2606.03347v1 Announce Type: cross Abstract: Score-based diffusion models have emerged as prominent deep generative models; however, their application to tabular data remains challenging because

applicationsarxiv-cs-ai

Bridging Auxiliary Constraints to Resolve Instruction Following in Large Reasoning Models

arXiv:2606.03624v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) have demonstrated impressive capabilities in many tasks, yet they struggle with reliably following multiple instructions,

ClinicalMC: A Benchmark for Multi-Course Clinical Decision-Making with Large Language Models

arXiv:2606.03157v1 Announce Type: new Abstract: Large language models (LLMs) have been widely adopted in healthcare, yet they still encounter significant challenges in complex clinical decision-making

Local Ai

Expert-Aware Causal Tracing of Factual Recall in Sparse MoE Language Models

arXiv:2606.03780v1 Announce Type: new Abstract: Causal tracing of factual recall has been studied predominantly in dense transformer language models, where interventions localize information flow to l

local-aiarxiv-cs-cl

Frontier models are powerful advisors. On @harvey's Legal Agent Benchmark, a GLM 5.1 worker using Claude Opus 4.7 as a sparse advisor reache…

model-releasesfireworks-ai--x

Frontier models are powerful advisors. On @harvey's Legal Agent Benchmark, a GLM 5.1 worker using Claude Opus 4.7 as a sparse advisor reached 18/100 all-pass versus 14/100 for Opus alone, at 39% of th

Large Language Models Are Overconfident in Their Own Responses

arXiv:2606.03437v1 Announce Type: new Abstract: Prior work has shown that instruction-tuned large language models (LLMs) are less well calibrated than their base pre-trained counterparts. However, lit

safetyarxiv-cs-cl

Measuring Weak-to-Strong Legibility of Reasoning Models

arXiv:2603.20508v2 Announce Type: replace-cross Abstract: Reasoning language models (RLMs) and the intermediate chains of thought they emit play an increasingly central role in multi-agent setups such

Non-Identical Diffusion Models in MIMO-OFDM Channel Generation

arXiv:2509.01641v3 Announce Type: replace-cross Abstract: We propose a novel diffusion model, termed the non-identical diffusion model, and investigate its application to wireless orthogonal frequency