Search: “pretraining”

arXiv:2605.10640v1 Announce Type: cross Abstract: Continual Pre-Training (CPT) is essential for enabling Language Models (LMs) to integrate new knowledge without erasing old. While classical CPT techn

tutorialsarxiv-cs-ai

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

TrajTok: Learning Trajectory Tokens enables better Video Understanding

DGX agent

arXiv:2602.22779v2 Announce Type: replace Abstract: Tokenization in video models, typically through patchification, generates an excessive and redundant number of tokens. This severely limits video ef

safetyarxiv-cs-cv

12 May 2026

Transfer Learning for Dead Fuel Moisture Prediction Using Time-Warping Recurrent Neural Networks

arXiv:2605.08379v1 Announce Type: cross Abstract: This paper proposes a time-warping transfer learning method, a technique for temporally rescaling the learned dynamics of a recurrent neural network (

Unified Noise Steering for Efficient Human-Guided VLA Adaptation

arXiv:2605.10821v1 Announce Type: new Abstract: Diffusion-based vision-language-action (VLA) models have emerged as strong priors for robotic manipulation, yet adapting them to real-world distribution

safetyarxiv-cs-ro

Agents

VECTOR-Drive: Tightly Coupled Vision-Language and Trajectory Expert Routing for End-to-End Autonomous Driving

arXiv:2605.08830v1 Announce Type: cross Abstract: End-to-end autonomous driving requires models to understand traffic scenes, infer driving intent, and generate executable motion plans. Recent vision-

agentsarxiv-cs-ai

VEGA: Visual Encoder Grounding Alignment for Spatially-Aware Vision-Language-Action Models

model-releasesarxiv-cs-ro

arXiv:2605.10485v1 Announce Type: new Abstract: Precise spatial reasoning is fundamental to robotic manipulation, yet the visual backbones of current vision-language-action (VLA) models are predominan

Why is prompting hard? Understanding prompts on binary sequence predictors

arXiv:2502.10760v2 Announce Type: replace Abstract: Frontier models can be prompted or conditioned to do many tasks, but finding good prompts is not always easy, nor is understanding some performant p

researcharxiv-cs-cl

WISTERIA: Learning Clinical Representations from Noisy Supervision via Multi-View Consistency in Electronic Health Records

arXiv:2605.09765v1 Announce Type: cross Abstract: Representation learning in electronic health records (EHR) has largely followed paradigms inherited from natural language processing, relying on seque

Yeti: A compact protein structure tokenizer for reconstruction and multi-modal generation

arXiv:2605.09981v1 Announce Type: cross Abstract: Multimodal models that jointly reason over protein sequences, structures, and function annotations within a unified representation hold immense potent

researcharxiv-cs-ai

YEZE at SemEval-2026 Task 9: Detecting Multilingual, Multicultural and Multievent Online Polarization via Heterogeneous Ensembling

arXiv:2605.06231v2 Announce Type: replace Abstract: This paper presents our system for SemEval-2026 Task 9: Detecting Multilingual, Multicultural and Multievent Online Polarization, which identifies p

researcharxiv-cs-cl

Anisotropic Modality Align

arXiv:2605.07825v1 Announce Type: cross Abstract: Training multimodal large language models has long been limited by the scarcity of high-quality paired multimodal data. Recent studies show that the s

safetyarxiv-cs-cv

Arrow: A Foundation Model for Causal Discovery

arXiv:2605.07204v1 Announce Type: new Abstract: We introduce Arrow, a foundation model for zero-shot causal discovery on observational tabular data. Arrow factorizes a directed acyclic graph into an u

Conservative Flows: A New Paradigm of Generative Models

arXiv:2605.06905v1 Announce Type: new Abstract: Modern generative modeling is dominated by transport from a noise prior to data. We propose an alternative paradigm in which generation is performed by

Consistency Regularised Gradient Flows for Inverse Problems

arXiv:2605.07907v1 Announce Type: cross Abstract: Vision-Language Latent Diffusion Models (LDMs) (Rombach et al., 2022) provide powerful generative priors for inverse problems. However, existing LDM-b

Convergence and Emergence of In-Context Reinforcement Learning with Chain of Thought

arXiv:2605.07123v1 Announce Type: new Abstract: In-context reinforcement learning (ICRL) refers to the ability of RL agents to adapt to new tasks at inference time without parameter updates by conditi

Coupling Models for One-Step Discrete Generation

arXiv:2605.07193v1 Announce Type: new Abstract: Generative modeling over discrete structures underpins applications across deep learning, from biological sequence design and code generation to large l

Dino U-Net: Exploiting High-Fidelity Dense Features from Foundation Models for Medical Image Segmentation

model-releasesarxiv-cs-cv

arXiv:2508.20909v2 Announce Type: replace Abstract: Foundation models pre-trained on large-scale natural image datasets offer a powerful paradigm for medical image segmentation. However, effectively t

Does Your Neural Network Extrapolate? Feature Engineering as Identifiability Bias for OOD Generalization

arXiv:2605.07483v1 Announce Type: cross Abstract: Successful deep neural networks discover salient features of data. We show when and why they fail to learn out-of-distribution (OOD)-relevant represen

Agents

Dynamic Mode Decomposition along Depth in Vision Transformers

arXiv:2605.07556v1 Announce Type: new Abstract: Recent work has shown that contiguous vision transformer (ViT) blocks (a) can be replaced by a linear map and (b) organize into recurrent phases of comp

agentsarxiv-cs-cv

EditTransfer++: Toward Faithful and Efficient Visual-Prompt-Guided Image Editing

arXiv:2605.07455v1 Announce Type: new Abstract: Visual-prompt-guided edit transfer aims to learn image transformations directly from example pairs, offering more precise and controllable editing than

safetyarxiv-cs-cv

Emergent Symbolic Structure in Health Foundation Models: Extraction, Alignment, and Cross-Modal Transfer

arXiv:2605.07407v1 Announce Type: new Abstract: Health foundation models (FMs) learn useful representations from wearable sensors, but interpreting what they encode and transferring that knowledge acr

safetyarxiv-cs-lg

Applications

Flexible Routing via Uncertainty Decomposition

arXiv:2605.07805v1 Announce Type: new Abstract: A key strategy for balancing performance and cost in modern machine learning systems is to dynamically route queries to either a low-cost model or a mor

applicationsarxiv-cs-lg

Graph Representation Learning Augmented Model Manipulation on Federated Fine-Tuning of LLMs

arXiv:2605.07961v1 Announce Type: new Abstract: Federated fine-tuning (FFT) has emerged as a privacy-preserving paradigm for collaboratively adapting large language models (LLMs). Built upon federated

Implicit Preference Alignment for Human Image Animation

model-releasesarxiv-cs-ai

arXiv:2605.07545v1 Announce Type: cross Abstract: Human image animation has witnessed significant advancements, yet generating high-fidelity hand motions remains a persistent challenge due to their hi

Local Ai

Instruction Tuning Changes How Upstream State Conditions Late Readout: A Cross-Patching Diagnostic

arXiv:2605.07284v1 Announce Type: new Abstract: Recent interpretability work has identified model-internal handles on post-trained behavior, including refusal directions, assistant/persona axes, and s

local-aiarxiv-cs-lg

Memory-Efficient Looped Transformer: Decoupling Compute from Memory in Looped Language Models

arXiv:2605.07721v1 Announce Type: cross Abstract: Recurrent LLM architectures have emerged as a promising approach for improving reasoning, as they enable multi-step computation in the embedding space

researcharxiv-cs-ai

Modality Gap-Driven Subspace Alignment Training Paradigm For Multimodal Large Language Models

arXiv:2602.07026v2 Announce Type: replace-cross Abstract: Despite the success of multimodal contrastive learning in aligning visual and linguistic representations, a persistent geometric anomaly, the

Normalizing Trajectory Models

arXiv:2605.08078v1 Announce Type: new Abstract: Diffusion-based models decompose sampling into many small Gaussian denoising steps -- an assumption that breaks down when generation is compressed to a

One Token Per Frame: Reconsidering Visual Bandwidth in World Models for VLA Policy

arXiv:2605.07931v1 Announce Type: cross Abstract: Vision-language-action (VLA) models increasingly rely on auxiliary world modules to plan over long horizons, yet how such modules should be parameteri

PET-Adapter: Test-Time Domain Adaptation for Full and Limited-Angle PET Image Reconstruction

arXiv:2605.08030v1 Announce Type: new Abstract: Positron Emission Tomography (PET) image reconstruction is inherently challenged by Poisson noise and physical degradation factors, which are further ex

Rethinking Dense Optical Flow without Test-Time Scaling

model-releasesarxiv-cs-cv

arXiv:2605.08000v1 Announce Type: new Abstract: Recent progress in dense optical flow has been driven by increasingly complex architectures and multi-step refinement for test-time scaling. While these

Agents

See Tomorrow, Act Today: Foresight-Driven Autonomous Driving

arXiv:2605.07195v1 Announce Type: new Abstract: Current end-to-end autonomous driving planners are fundamentally reactive: they condition on historical and present observations to predict future actio

agentsarxiv-cs-cv

Sign-Based Optimizers Are Effective Under Heavy-Tailed Noise

arXiv:2602.07425v2 Announce Type: replace-cross Abstract: While adaptive gradient methods are the workhorse of modern machine learning, sign-based optimization algorithms such as Lion and Muon have re

researcharxiv-cs-cl

Sparse Autoencoders as Plug-and-Play Firewalls for Adversarial Attack Detection in VLMs

arXiv:2605.07447v1 Announce Type: cross Abstract: Vision-language models (VLMs) have advanced rapidly and are increasingly deployed in real-world applications, especially with the rise of agent-based

Star Elastic: Many-in-One Reasoning LLMs with Efficient Budget Control

arXiv:2605.07182v1 Announce Type: new Abstract: Training a family of large language models (LLMs), either from scratch or via iterative compression, is prohibitively expensive and inefficient, requiri

STARFlow2: Bridging Language Models and Normalizing Flows for Unified Multimodal Generation

arXiv:2605.08029v1 Announce Type: new Abstract: Deep generative models have advanced rapidly across text and vision, motivating unified multimodal systems that can understand, reason over, and generat

Stochastic Transition-Map Distillation for Fast Probabilistic Inference

arXiv:2605.07661v1 Announce Type: cross Abstract: Diffusion models achieve strong generation quality, diversity, and distribution coverage, but their performance often comes with expensive inference.

Structured Prototype-Guided Adaptation for EEG Foundation Models

arXiv:2602.17251v2 Announce Type: replace Abstract: Electroencephalography (EEG) foundation models (EFMs) have shown strong potential for transferable representation learning, yet their adaptation in

Local Ai

Test-Time Compositional Generalization in Diffusion Models via Concept Discovery

arXiv:2605.07078v1 Announce Type: new Abstract: Compositional generalization requires models to produce novel configurations from familiar parts. In diffusion models, prior compositional generation me

local-aiarxiv-cs-lg

The Convergence Gap: Instruction-Tuned Language Models Stabilize Later in the Forward Pass

arXiv:2605.07282v1 Announce Type: new Abstract: Final outputs hide when a checkpoint commits to its next-token prediction. We introduce the convergence gap, a model-diffing diagnostic that decodes eac

The Position Curse: LLMs Struggle to Locate the Last Few Items in a List

model-releasesarxiv-cs-cl

arXiv:2605.07127v1 Announce Type: cross Abstract: Modern large language models (LLMs) can find a needle in a haystack (locating a single relevant fact buried among hundreds of thousands of irrelevant

Tool Calling is Linearly Readable and Steerable in Language Models

model-releasesarxiv-cs-ai

arXiv:2605.07990v1 Announce Type: cross Abstract: When a tool-calling agent picks the wrong tool, the failure is invisible until execution: the email gets sent, the meeting gets missed. Probing 12 ins

Toward Privileged Foundation Models:LUPI for Accelerated and Improved Learning

arXiv:2605.07799v1 Announce Type: cross Abstract: Training foundation models is computationally intensive and often slow to converge.We introduce PIQL,Privileged Information for Quick and Quality Lear

researcharxiv-cs-ai

Towards Explainable Industrial Anomaly Detection via Knowledge-Guided Latent Reasoning

arXiv:2602.09850v2 Announce Type: replace Abstract: Industrial anomaly detection demands precise reasoning over fine-grained defect patterns. However, existing multimodal large language models (MLLMs)

Local Ai

Unlocking High-Fidelity Molecular Generation from Mass Spectra via Dual-Stream Line Graph Diffusion

arXiv:2605.07048v1 Announce Type: cross Abstract: De novo molecular generation from tandem mass spectra is a challenging inverse problem whose core difficulty lies in the circular dependency between a

local-aiarxiv-cs-ai

VLA-GSE: Boosting Parameter-Efficient Fine-Tuning in VLA with Generalized and Specialized Experts

model-releasesarxiv-cs-ro

arXiv:2605.06175v2 Announce Type: replace Abstract: Vision-language-action (VLA) models inherit rich visual-semantic priors from pre-trained vision-language backbones, but adapting them to robotic con

What Matters for Diffusion-Friendly Latent Manifold? Prior-Aligned Autoencoders for Latent Diffusion

arXiv:2605.07915v1 Announce Type: new Abstract: Tokenizers are a crucial component of latent diffusion models, as they define the latent space in which diffusion models operate. However, existing toke

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…