Search: “pretraining”

arXiv:2605.20342v1 Announce Type: new Abstract: Training large multimodal models (LMMs) via reinforcement learning (RL) to natively invoke video-processing tools (e.g., cropping) has become a promisin

agentsarxiv-cs-cv

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

Rethinking Cross-Layer Information Routing in Diffusion Transformers

DGX agent

arXiv:2605.20708v1 Announce Type: new Abstract: Diffusion Transformers (DiTs) have become a de facto backbone of modern visual generation, and nearly every major axis of their design -- tokenization,

safetyarxiv-cs-cv

21 May 2026

Local Ai

Semantic Granularity Navigation in Image Editing

arXiv:2605.21190v1 Announce Type: new Abstract: Despite the generative capabilities of diffusion and flow models, real-image editing remains constrained by a persistent trade-off between semantic edit

local-aiarxiv-cs-cv

SHINE: A Scalable In-Context Hypernetwork for Mapping Context to LoRA in a Single Pass

model-releasesarxiv-cs-cl

arXiv:2602.06358v2 Announce Type: replace Abstract: We propose SHINE (Scalable Hyper In-context NEtwork), a scalable hypernetwork that can map diverse meaningful contexts into high-quality LoRA adapte

SMoA: Spectrum Modulation Adapter for Parameter-Efficient Fine-Tuning

model-releasesarxiv-cs-cl

arXiv:2605.21147v1 Announce Type: cross Abstract: As the number of model parameters increases, parameter-efficient fine-tuning (PEFT) has become the go-to choice for tailoring pre-trained large langua

Spectral Unforgetting: Post-Hoc Recovery of Damaged Capabilities Without Retraining

model-releasesarxiv-cs-lg

arXiv:2605.20296v1 Announce Type: new Abstract: Fine-tuning a language model for a target task routinely degrades capabilities the training data never explicitly threatened. We study this phenomenon,

STEAM: A Training-Free Congestion-Aware Enhancement Framework for Decentralized Multi-Agent Path Finding

arXiv:2605.20929v1 Announce Type: new Abstract: We propose STEAM (Spatial, Temporal, and Emergent congestion Awareness for MAPF), a training-free test-time enhancement framework for learning-based dec

safetyarxiv-cs-ro

Synchronization and Turn-Taking in Full-Duplex Speech Dialogue Models

arXiv:2605.20356v1 Announce Type: new Abstract: Full-duplex spoken dialogue models (SDMs) can listen and speak simultaneously, enabling interaction dynamics closer to human conversation than turn-base

safetyarxiv-cs-cl

TriForces: Augmenting Atomistic GNNs for Transferable Representations

arXiv:2605.20581v1 Announce Type: new Abstract: Machine learning interatomic potentials (MLIPs) achieve excellent accuracy when trained on large Density Functional Theory (DFT) data. To be useful in p

VISTA: Technical Report for the Ego4D Short-Term Object Interaction Anticipation at EgoVis 2026

arXiv:2605.20901v1 Announce Type: new Abstract: We propose VISTA, a V-JEPA Integrated StillFast Temporal Anticipator for the Ego4D Short-Term Object Interaction Anticipation (STA) Challenge at EgoVis

WestWorld: A Knowledge-Encoded Scalable Trajectory World Model for Diverse Robotic Systems

arXiv:2603.14392v2 Announce Type: replace Abstract: Trajectory world models play a crucial role in robotic dynamics learning, planning, and control. While recent works have explored trajectory world m

applicationsarxiv-cs-lg

WildRoadBench: A Wild Aerial Road-Damage Grounding Benchmark for Vision-Language Models and Autonomous Agents

arXiv:2605.20306v1 Announce Type: new Abstract: We introduce WildRoadBench, a wild aerial road-damage grounding benchmark that couples direct visual grounding by vision-language models with autonomous

Awakening the Hydra: Stabilizing Multi-Concept Backdoor Injection in Text-to-Image Diffusion Models

arXiv:2605.19698v1 Announce Type: cross Abstract: Text-to-image diffusion models are increasingly developed through open-source reuse and repeated downstream fine-tuning, where reused checkpoints are

Beyond Prediction Accuracy: Target-Space Recovery Profiles for Evaluating Model-Brain Alignment

arXiv:2605.20127v1 Announce Type: cross Abstract: Artificial vision models are often evaluated against the human visual cortex by measuring how accurately their internal representations predict brain

Tutorials

Character-Centered Dialogue Generation from Scene-Level Prompts

arXiv:2505.16819v4 Announce Type: replace Abstract: Recent advances in scene-based video generation enable coherent visual narratives from structured prompts, yet a key aspect of storytelling -- chara

tutorialsarxiv-cs-cv

CounterFlow: A Two-Phase Inference-Time Sampling for Counterfactual Video Foley Generation

arXiv:2605.18916v1 Announce Type: cross Abstract: We investigate Counterfactual Video Foley Generation, which aims to adopt a sound-source identity that contradicts the visual evidence while remaining

D^3-Subsidy: Online and Sequential Driver Subsidy Decision-Making for Large-Scale Ride-Hailing Market

model-releasesarxiv-cs-lg

arXiv:2605.20036v1 Announce Type: new Abstract: Ride-hailing platforms like DiDi Chuxing operate in highly dynamic environments where balancing driver supply and passenger demand is critical. Although

Data-Efficient Self-Supervised Algorithms for Fine-Grained Birdsong Analysis

arXiv:2511.12158v3 Announce Type: replace Abstract: Research in bioacoustics, neuroscience, and linguistics often uses birdsong as a proxy to acquire knowledge across diverse areas. This requires audi

applicationsarxiv-cs-lg

Delta Attention Residuals

arXiv:2605.18855v1 Announce Type: cross Abstract: Attention Residuals replace standard additive residual connections with learned softmax attention over previous layer outputs, enabling selective cros

Distributional Energy-Based Models for Uncertainty-Aware Structured LLM Reasoning

arXiv:2605.18871v1 Announce Type: cross Abstract: When Large Language Models produce structured outputs such as travel plans, code solutions, or multi-step proofs, individual reasoning steps may appea

Dr.LLM: Dynamic Layer Routing in LLMs

arXiv:2510.12773v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) process every token through all layers of a transformer stack, causing wasted computation on simple queries and i

Dual-Prompt CLIP with Hybrid Visual Encoders for Occluded Person Re-Identification

arXiv:2605.19527v1 Announce Type: new Abstract: Occluded person re-identification focuses on matching partially visible pedestrians across multiple camera views. However, occlusions disrupt body-regio

DynaSTy: A Framework for SpatioTemporal Node Attribute Prediction in Dynamic Graphs

arXiv:2601.05391v2 Announce Type: replace Abstract: Accurate multistep forecasting of node-level attributes on dynamic graphs is critical for applications ranging from financial trust networks to biol

safetyarxiv-cs-lg

Fast Tensorization of Neural Networks via Slice-wise Feature Distillation

arXiv:2605.19842v1 Announce Type: new Abstract: We propose a scalable tensorization framework for neural network compression based on slice-wise feature distillation. Unlike conventional tensor decomp

FlexDraft: Flexible Speculative Decoding via Attention Tuning and Bonus-Guided Calibration

arXiv:2605.20022v1 Announce Type: new Abstract: Speculative decoding accelerates memory-bound LLM inference without quality degradation by using a fast drafter to propose multiple candidate tokens and

researcharxiv-cs-cl

HELLoRA: Hot Experts Layer-Level Low-Rank Adaptation for Mixture-of-Experts Models

arXiv:2605.18795v1 Announce Type: cross Abstract: Low-Rank Adaptation (LoRA) dominates parameter-efficient fine-tuning of large language models, yet most variants target dense architectures. Mixture-o

LLM Benchmark Datasets Should Be Contamination-Resistant

arXiv:2605.19999v1 Announce Type: cross Abstract: Benchmark datasets are critical for reproducible, reliable, and discriminative evaluation of LLMs. However, recent studies reveal that many benchmark

LoRA vs. Full Fine-Tuning: A Theoretical Perspective

arXiv:2605.19018v1 Announce Type: new Abstract: Fine-tuning adapts a pre-trained model to downstream tasks using a small amount of labeled data. Low-Rank Adaptation (LoRA) is an efficient fine-tuning

Mathematical Reasoning in Large Language Models: Benchmarks, Architectures, Evaluation, and Open Challenges

arXiv:2605.19723v1 Announce Type: cross Abstract: Mathematical reasoning is essential for problem-solving in education, science, and industry, serving as a crucial benchmark for evaluating artificial

Hardware

Prior Knowledge or Search? A Study of LLM Agents in Hardware-Aware Code Optimization

arXiv:2605.19782v1 Announce Type: new Abstract: LLM discovery and optimization systems are increasingly applied across domains, implementing a common propose-evaluate-revise loop. Such optimization or

hardwarearxiv-cs-ai

Prognostic Value of Lung Ultrasound Biomarkers for Readmission Risk in Congestive Heart Failure: A Pilot Data-Driven Analysis

arXiv:2605.18878v1 Announce Type: cross Abstract: Hospital readmission within 30 days of discharge is a leading driver of morbidity, mortality, and avoidable healthcare expenditure in congestive heart

applicationsarxiv-cs-cv

Rapid patient-specific neural networks for intraoperative X-ray to volume registration

arXiv:2503.16309v2 Announce Type: replace-cross Abstract: Advanced navigation techniques in image-guided interventions and surgical robotics require the rapid and precise alignment of 3D preoperative

safetyarxiv-cs-cv

Real-World On-Vehicle Evaluation of Embedding-Based Anomaly Detection

arXiv:2605.19744v1 Announce Type: new Abstract: Detecting anomalies in traffic scenes is crucial for ensuring safety in autonomous driving, yet collecting representative anomalous data remains challen

ScheduleFree+: Scaling Learning-Rate-Free & Schedule-Free Learning to Large Language Models

arXiv:2605.19095v1 Announce Type: cross Abstract: Schedule-Free Learning has shown promise as a practical anytime training method for machine learning, showing success across dozens of standard benchm

Shaping the Prior: How Synthetic Task Distributions Determine Tabular Foundation Model Quality

arXiv:2605.18971v1 Announce Type: cross Abstract: What determines the quality of a tabular foundation model? Unlike language or vision, tabular foundation models acquire their inductive biases almost

Tutorials

Smartphone-based Circular Plot Sampling for Forest Inventory

arXiv:2605.19213v1 Announce Type: new Abstract: Circular sample plots are a cornerstone of forest inventory, yet accurate measurement of tree diameter at breast height (DBH) and spatial location withi

tutorialsarxiv-cs-cv

Spark3R: Asymmetric Token Reduction Makes Fast Feed-Forward 3D Reconstruction

arXiv:2605.06270v2 Announce Type: replace Abstract: Feed-forward 3D reconstruction models based on Vision Transformers can directly estimate scene geometry and camera poses from a small set of input i

Sparse Mixture-of-Experts Routing in Visual Diffusion Transformers:Diagnosis, Boundary Calibration and Evolutionary Roadmap from Routing Collapse to Selective Deadlock

arXiv:2605.19378v1 Announce Type: new Abstract: This paper systematically diagnoses the training failure modes of Token-Choice sparse Mixture-of-Experts (MoE) on video Diffusion Transformers. Starting

Spatial-MLLM: Boosting MLLM Capabilities in Visual-based Spatial Intelligence

arXiv:2505.23747v2 Announce Type: replace-cross Abstract: Recent advancements in Multimodal Large Language Models (MLLMs) have significantly enhanced performance on 2D visual tasks. However, improving

applicationsarxiv-cs-ai

SpecX: A Large-Scale Benchmark for Multi-Modal Spectroscopy and Cross-Paradigm Evaluation

arXiv:2605.18791v1 Announce Type: cross Abstract: Existing spectral benchmarks are limited in scale, modality alignment, and evaluation scope, and typically focus on either specialized models or multi

Local Ai

Tango3D: Towards Alignment for Global and Local 2D-3D Correspondence

arXiv:2605.19727v1 Announce Type: new Abstract: Existing 3D foundation models typically align point clouds to frozen vision-language spaces like CLIP, which achieve strong cross-modal retrieval by com

local-aiarxiv-cs-cv

The Routing and Filtering Structure of Attention

arXiv:2605.18826v1 Announce Type: cross Abstract: The attention interaction matrix QK^{op} contains two entangled computations: a skew-symmetric component that redistributes information between positi

VCR: Learning Valid Contextual Representation for Incomplete Wearable Signals

arXiv:2605.18837v1 Announce Type: cross Abstract: Wearable devices enable continuous health monitoring from multimodal signals, but real-world deployment is hindered by limited labeled data and pervas

applicationsarxiv-cs-ai

What Really Improves Mathematical Reasoning: Structured Reasoning Signals Beyond Pure Code

arXiv:2605.19762v1 Announce Type: new Abstract: Code has become a standard component of modern foundation language model (LM) training, yet its role beyond programming remains unclear. We revisit the

1GC-7RC: One Graphic Card -- Seven Research Challenges! How Good Are AI Agents at Doing Your Job?

arXiv:2605.17046v1 Announce Type: cross Abstract: Autonomous AI coding agents are becoming a core tool for ML practitioners in industry and research alike. Despite this growing adoption, no standardiz

A Large-Scale Study on the Accuracy vs Cost Trade-offs of Training and Evaluation Settings in Fine-Grained Image Recognition

arXiv:2605.18700v1 Announce Type: new Abstract: Prior work on fine-grained image recognition (FGIR) has established the importance of the backbone selection, but has neglected the accuracy-vs-cost tra

Adaptive Fused Prior Transfer for Controllable Generative Image Compression

arXiv:2605.16817v1 Announce Type: cross Abstract: Learned image compression has achieved competitive rate-distortion performance, but very-low-bitrate reconstruction remains difficult because the tran

Local Ai

AtlasVid: Efficient Ultra-High-Resolution Long Video Generation via Decoupled Global-Local Modeling

arXiv:2605.16649v1 Announce Type: new Abstract: Recent diffusion-based video generators have achieved remarkable visual fidelity and prompt controllability, yet scaling them to ultra-high-resolution (

local-aiarxiv-cs-cv