Search: “research”

Simulation-Free Estimation of Traffic Flows from Sparse Count Data

arXiv:2606.23536v1 Announce Type: new Abstract: We propose a method for estimating time-varying traffic flow patterns from sparse aggregated vehicle counts. The method partitions the study area into s

SLeDGe: Semi-Supervised Learning on Data Streams with Graph Structure Learning

DGX agent

arXiv:2606.21096v1 Announce Type: new Abstract: Semi-supervised learning (SSL) on data streams is challenging due to the continuous evolution of high-volume data and the scarcity of labels. Existing m

researcharxiv-cs-lg

23 Jun 2026

SOAP-Bubbles: Structured Weight Uncertainty for Neural Networks

arXiv:2606.23357v1 Announce Type: new Abstract: Structured weight-uncertainty can improve many aspects of deep learning, but it remains costly to estimate and difficult to implement. Here, we show tha

Spatially Grounded Concept-Based Image Classification

arXiv:2510.04180v2 Announce Type: replace Abstract: Deep neural networks can achieve high accuracy while relying on evidence that is hard to inspect or misaligned with the intended task. Concept Bottl

Specificity- and Calibration-Aware Breast Ultrasound Segmentation via Entropy-Guided Boundary Supervision

arXiv:2606.22308v1 Announce Type: cross Abstract: Lesion segmentation in breast ultrasound involves two related challenges. In images with lesions, speckle noise, low tissue contrast, and posterior ac

Spectrally Safe Neural Operator Warm-Starts for Large-Scale Newton Solvers

arXiv:2606.21828v1 Announce Type: cross Abstract: Neural operators are increasingly used to warm-start Newton solvers for nonlinear PDEs, on the premise that a low test error places the initial guess

SpotAttention: Plug-In Block-Sparse Routing for Pretrained Long-Context Transformers

arXiv:2606.22874v1 Announce Type: new Abstract: Long contexts have become standard in pretrained LLMs, yet they remain expensive to run: prefill compute grows quadratically with sequence length, and e

Stage-dependent integer-binary encoding in factorization-machine black-box optimization

arXiv:2606.23188v1 Announce Type: new Abstract: Black-box optimization (BBO) deals with problems where objective functions lack explicit analytical forms and are expensive to evaluate. Factorization m

Stealthy World Model Manipulation via Data Poisoning

arXiv:2606.18697v2 Announce Type: replace Abstract: Model-based learning agents use learned world models to predict future states, plan actions, and adapt to new environments. However, the process of

Stochastic Signed Distance Processes

arXiv:2606.20856v1 Announce Type: new Abstract: Multi-view surface reconstruction is a core problem in computer vision. One prominent line of work represents the surface implicitly as a signed distanc

StreamPPG: Low-Latency rPPG Estimation via Consistent Privileged Learning

arXiv:2606.23186v1 Announce Type: new Abstract: Remote photoplethysmography (rPPG) estimates the blood volume pulse (BVP) signal from facial videos, enabling contact-free health monitoring. Convention

StructSAM: Structure- and Spectrum-Preserving Token Merging for Segment Anything Models

arXiv:2603.07307v2 Announce Type: replace Abstract: Recent token merging techniques for Vision Transformers (ViTs) provide substantial speedups by reducing the number of tokens processed by self-atten

Structure-Aware Compound-Protein Affinity Prediction via Graph Neural Networks with Group Lasso Regularization

arXiv:2507.03318v3 Announce Type: replace Abstract: Explainable artificial intelligence approaches accelerate drug discovery by improving molecular representation learning, identifying key molecular s

Structured Spectral Graph Representation Learning for Multi-label Abnormality Analysis from 3D CT Scans

arXiv:2510.10779v5 Announce Type: replace Abstract: With the growing volume of CT examinations, there is an increasing demand for automated tools such as organ segmentation, abnormality detection, and

Sublinearly Structured Deep Neural Networks Achieve Feature Learning Consistency for Compositional Functions

arXiv:2606.23477v1 Announce Type: cross Abstract: Over the past decade, deep neural networks (DNNs) have achieved remarkable success on complex machine-learning tasks, yet the theoretical foundations

Super Mario is mathier than you think

Here’s a problem you probably didn’t solve in school: You’re an ambitious young plumber from Brooklyn in a world inhabited by violent human-size mushrooms called Goombas. The love of your life has bee

researchmit-tech-review

Surgical Anatomy Recognition with Context Learning using Foundation Representations

arXiv:2606.22124v1 Announce Type: new Abstract: Accurate recognition of anatomical structures is essential for safe and effective minimally invasive surgery (MIS), yet it remains underexplored in surg

Synthetic Network Packet Generation through Statistical Learning and Genetic Algorithms

arXiv:2606.20864v1 Announce Type: cross Abstract: Developing robust intrusion detection systems (IDS) for IoT environments requires large, labeled datasets capturing realistic traffic distributions ac

T-MOR: Learning Motion-Aware Skeleton Representations for Human Action Recognition

arXiv:2606.21607v1 Announce Type: new Abstract: Vision-language models such as CLIP have recently achieved strong performance on a wide range of visual understanding tasks. However, most existing mode

T-Rex: Tactile-Reactive Dexterous Manipulation

arXiv:2606.17055v2 Announce Type: replace Abstract: The ability to react dynamically to tactile signals has long been considered crucial to agile human-level dexterity. Yet contemporary learning-based

researcharxiv-cs-ro

T-VSS: Test-Time Visual Subspace Steering for Adversarial Robustness of Vision-Language Models

arXiv:2606.23132v1 Announce Type: new Abstract: Vision-language models (VLMs) achieve strong zero-shot recognition, but they remain highly vulnerable to adversarial perturbations. Recent test-time ada

Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection

arXiv:2509.21606v3 Announce Type: replace Abstract: Federated continual learning (FCL) enables collaborative model training across distributed clients on sequentially arriving tasks without revisiting

Tensor Train Decomposition-based 3D Implicit Full Waveform Inversion with Multi-scale Structural Similarity

arXiv:2606.22867v1 Announce Type: cross Abstract: Three-dimensional full waveform inversion (3DFWI) is a powerful technique for reconstructing high-resolution subsurface velocity models. However, its

TF-SNO: Time-Frequency Gated Spectral Neural Operators for Learning Non-Stationary Partial Differential Equations

arXiv:2606.21189v1 Announce Type: new Abstract: Non-stationary partial differential equations (PDEs) arise throughout scientific computing, where the dominant frequency content and energy distribution

The $400 million machine powering the future of chipmaking

Jos Benschop is climbing a ladder to get to the top of his newest machine. It’s a bit of a schlep. The contraption is the size of a double-decker bus—more than 150 tons of gleaming precision-milled al

researchmit-tech-review

The Anatomy of the CTC Oracle Gap: Acoustic Exhaustion and Linguistic Recovery

arXiv:2606.23306v1 Announce Type: cross Abstract: We study the limits of CTC-internal scoring for N-best hypothesis selection and locate the information bottleneck separating acoustic confidence from

The Cost Geometry of Belief: finite-resource inference under noisy observation

arXiv:2606.21585v1 Announce Type: new Abstract: We equip the space of beliefs with a cost geometry (what it costs to pass from one belief to another): optimal transport in Wasserstein space, reweighte

The Download: the future of chipmaking and Anthropic’s government clash

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. The $400 million machine powering the future of chipmaking It’

researchmit-tech-review

The Energy Consumption of Transformer Fine-Tuning: A Roofline-Inspired Scaling Model

arXiv:2606.23546v1 Announce Type: new Abstract: Transformer-based models underpin modern natural language processing but incur rapidly growing computational and energy costs. As training scales in bot

The First Assessment of PhiSat-2 Imagery for Monocular Building Height Estimation

arXiv:2603.29245v3 Announce Type: replace Abstract: Monocular building height estimation from optical imagery is important for characterizing urban vertical structure, yet remains challenging due to t

The Keras team is gonna be at the AI Systems DevLabs in Sunnvale tomorrow. Come say hi if you're around!

researchfrancois-chollet--x

François Chollet announced that members of the Keras team would be attending the AI Systems DevLabs event in Sunnyvale and invited people to meet them there. This post indicates a scheduled appearance

The Optimal Token Baseline: Variance Reduction for Long-Horizon LLM-RL

arXiv:2602.07078v2 Announce Type: replace Abstract: Reinforcement Learning (RL) for Large Language Models (LLMs) often suffers from training collapse in long-horizon tasks due to exploding gradient va

The Optimization Landscape of Caratheodory Decomposition of Toeplitz Covariances

arXiv:2511.01605v2 Announce Type: replace Abstract: Toeplitz covariance estimation is a classical problem in statistical signal processing, yet the geometry of the Gaussian maximum-likelihood objectiv

The Trilemma of Truth in Large Language Models

arXiv:2506.23921v5 Announce Type: replace-cross Abstract: The public often attributes human-like qualities to large language models (LLMs), assuming that they 'know' certain things. In reality, LLMs e

The Unreasonable Effectiveness of VLMs for Zero-shot Procedural Mistake Detection

arXiv:2606.21579v1 Announce Type: new Abstract: Procedural mistake detection is important for quality control and user assistance across many disciplines. Recent work in this field has achieved signif

ThermoLLM: Thermodynamics-Aware HVAC Control with Spatial-Semantic Knowledge Graph

arXiv:2606.22911v1 Announce Type: cross Abstract: Multi-zone HVAC control is a spatial decision problem in which indoor thermal evolution and control decisions depend not only on outdoor conditions an

TIDY: Thermal Infrared Image Denoising via Wavelet Domain Entropy and Directional Stripe Index

arXiv:2606.19813v1 Announce Type: cross Abstract: Thermal infrared (TIR) imaging has been a popular choice for field robotics due to its robust perception capability under low light visual degradation

TooBad: Backdoor Diffusion Models with Ultra-Low Poison Rate and Imperceptible Trigger

arXiv:2606.23362v1 Announce Type: cross Abstract: Diffusion models (DMs), despite their impressive capabilities across a wide range of generative tasks, have been shown to be vulnerable to backdoor at

Towards a Novel Wearable Robotic Vest for Hemorrhage Suppression

arXiv:2602.01448v2 Announce Type: replace Abstract: This paper introduces a novel robotic system designed to manage severe bleeding in emergency scenarios, including unique environments like space sta

researcharxiv-cs-ro

Towards Accurate and Robust Surveillance Roadside IVD via Trackletized Audio-Visual Reasoning

arXiv:2606.22299v1 Announce Type: new Abstract: Idling Vehicle Detection (IVD) seeks to determine, at the final frame of a video clip, whether any vehicle is idling, meaning the vehicle is stationary

Towards CSI-Native Foundation Models: A Channel-Adaptive Roadmap for 6G

arXiv:2606.20670v1 Announce Type: new Abstract: Wireless foundation models offer a path toward reusable channel state information (CSI) intelligence for sixth-generation (6G) systems. However, existin

Towards Robust Training in NNGPT AutoML Pipeline: A Loss-Optimizer Pairing Selection Study

arXiv:2606.20933v1 Announce Type: new Abstract: The choice of loss function and optimizer is an important decision, that shapes further model training. Yet automated architecture search pipelines (Aut

Towards Understanding the Power and Limits of the Muon Optimizer: A River-Valley Perspective

arXiv:2606.21514v1 Announce Type: new Abstract: Recently, Muon has gained substantial attention as an appealing alternative to Adam-like optimizers, with many works highlighting its advantages through

TraceMark-LDM: Authenticatable Watermarking for Latent Diffusion Models via Binary-Guided Rearrangement

arXiv:2503.23332v2 Announce Type: replace Abstract: Image generation algorithms are increasingly integral to diverse aspects of human society, driven by their practical applications. However, insuffic

TriMotion: Modality-Agnostic Camera Control for Video Generation

arXiv:2606.20774v1 Announce Type: new Abstract: Camera motion control is essential for directing viewpoint changes in generative systems. However, existing methods typically condition the generation p

Trustworthy MRI Reconstruction via Bayesian Uncertainty Quantification with Sparsity Prior Models

arXiv:2606.17343v2 Announce Type: replace Abstract: We propose a novel Bayesian framework for joint image reconstruction and uncertainty quantification from compressed sensing magnetic resonance imagi

Turning Tabular Foundation Models into Graph Foundation Models

arXiv:2508.20906v3 Announce Type: replace Abstract: While foundation models have revolutionized fields such as natural language processing and computer vision, their potential in graph machine learnin

Model Releases

Two-Bridge: Exclusive Objectives and Extended Horizon StarCraft II Benchmark

model-releasesarxiv-cs-lg

arXiv:2603.06608v2 Announce Type: replace-cross Abstract: The research community lacks a middle ground between StarCraft II full game and its mini-games. The full-game's sprawling state-action space r