research

Don't Pass@k: A Bayesian Framework for Large Language Model Evaluation

arXiv:2510.04265v4 Announce Type: replace-cross Abstract: Pass@k is widely used to report the reasoning performance of LLMs, but it often produces unstable and potentially misleading rankings, especia

researcharxiv-cs-cl

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

DP-{lambda}CGD: Efficient Noise Correlation for Differentially Private Model Training

DGX agent

arXiv:2601.22334v2 Announce Type: replace Abstract: Differentially private stochastic gradient descent (DP-SGD) is the gold standard for training machine learning models with formal differential priva

researcharxiv-cs-lg

13 May 2026

DriftXpress: Faster Drifting Models via Projected RKHS Fields

arXiv:2605.12183v1 Announce Type: new Abstract: Drifting Models have emerged as a new paradigm for one-step generative modeling, achieving strong image quality without iterative inference. The premise

DSA-NRP: No-Reflow Prediction from Angiographic Perfusion Dynamics in Stroke EVT

arXiv:2506.17501v3 Announce Type: replace-cross Abstract: Following successful large-vessel recanalization via endovascular thrombectomy (EVT) for acute ischemic stroke (AIS), some patients experience

EDGER: EDge-Guided with HEatmap Refinement for Generalizable Image Forgery Localization

arXiv:2605.12002v1 Announce Type: new Abstract: Text-guided inpainting has made image forgery increasingly realistic, challenging both SID and IFL. However, existing methods often struggle to point ou

Efficient Adjoint Matching for Fine-tuning Diffusion Models

arXiv:2605.11480v1 Announce Type: new Abstract: Reward fine-tuning has become a common approach for aligning pretrained diffusion and flow models with human preferences in text-to-image generation. Am

Efficient Bayesian Inference from Noisy Pairwise Comparisons

arXiv:2510.09333v2 Announce Type: replace-cross Abstract: Evaluating generative models is challenging because standard metrics often fail to reflect human preferences. Human evaluations are more relia

EgoForce: Forearm-Guided Camera-Space 3D Hand Pose from a Monocular Egocentric Camera

arXiv:2605.12498v1 Announce Type: new Abstract: Reconstructing the absolute 3D pose and shape of the hands from the user's viewpoint using a single head-mounted camera is crucial for practical egocent

Elastic Attention Cores for Scalable Vision Transformers

arXiv:2605.12491v1 Announce Type: new Abstract: Vision Transformers (ViTs) achieve strong data-driven scaling by leveraging all-to-all self-attention. However, this flexibility incurs a computational

Elicitation-Augmented Bayesian Optimization

arXiv:2605.12079v1 Announce Type: new Abstract: Human-in-the-loop Bayesian optimization (HITL BO) methods utilize human expertise to improve the sample-efficiency of BO. Most HITL BO methods assume th

EndoVGGT: GNN-Enhanced Depth Estimation for Surgical 3D Reconstruction

arXiv:2603.24577v2 Announce Type: replace Abstract: Accurate 3D reconstruction of deformable soft tissues is essential for surgical robotic perception. However, low-texture surfaces, specular highligh

Enriching and Controlling Global Semantics for Text Summarization

arXiv:2109.10616v2 Announce Type: replace Abstract: Recently, Transformer-based models have been proven effective in the abstractive summarization task by creating fluent and informative summaries. Ne

researcharxiv-cs-cl

EpiCastBench: Datasets and Benchmarks for Multivariate Epidemic Forecasting

arXiv:2605.11598v1 Announce Type: new Abstract: The increasing adoption of data-driven decision-making in public health has established epidemic forecasting as a critical area of research. Recent adva

EqOD: Symmetry-Informed Stability Selection for PDE Identification

arXiv:2605.11524v1 Announce Type: new Abstract: Data-driven identification of partial differential equations (PDEs) relies on sparse regression over a candidate library of differential operators, wher

Error whitening: Why Gauss-Newton outperforms Newton

arXiv:2605.11316v1 Announce Type: new Abstract: The Gauss-Newton matrix is widely viewed as a positive semidefinite approximation of the Hessian, yet mounting empirical evidence shows that Gauss-Newto

Evolutionary Task Discovery: Advancing Reasoning Frontiers via Skill Composition and Complexity Scaling

arXiv:2605.11666v1 Announce Type: new Abstract: The reasoning frontier of Large Language Models (LLMs) has advanced significantly through modern post-training paradigms (e.g., Reinforcement Learning f

Exploring Token-Space Manipulation in Latent Audio Tokenizers

arXiv:2605.11192v1 Announce Type: cross Abstract: Neural audio codecs provide compact discrete representations for speech generation and manipulation. However, most codecs organize tokens as frame-lev

Express Your Doubts -- Probabilistic World Modeling Should not be Based on Token logprobs

arXiv:2505.02072v2 Announce Type: replace Abstract: Language modeling has shifted in recent years from a distribution over strings to prediction models with textual inputs and outputs for general-purp

researcharxiv-cs-cl

Feasible Force Set Shaping for a Payload-Carrying Platform Consisting of Tiltable Multiple UAVs Connected Via Passive Hinge Joints

arXiv:2503.00341v2 Announce Type: replace Abstract: This paper presents a method for shaping the feasible force set of a payload-carrying platform composed of multiple Unmanned Aerial Vehicles (UAVs)

researcharxiv-cs-ro

Federated Client Selection under Partial Visibility: A POMDP Approach with Spatio-Temporal Attention

arXiv:2605.11752v1 Announce Type: new Abstract: Federated learning relies on effective client selection to alleviate the performance degradation caused by data heterogeneity. Most existing methods ass

Finite Sentence-Interface Control for Learning Bounded-Fan-Out Linear MCFGs under Fixed Monoid Typing

arXiv:2605.11644v1 Announce Type: cross Abstract: We study positive-data learning of bounded-fan-out linear multiple context-free grammars under a fixed explicit finite monoid homomorphism (h). The ma

FIS-DiT: Breaking the Few-Step Video Inference Barrier via Training-Free Frame Interleaved Sparsity

arXiv:2605.11869v1 Announce Type: new Abstract: While the overall inference latency of Video Diffusion Transformers (DiTs) can be substantially reduced through model distillation, per-step inference l

From Model Uncertainty to Human Attention: Localization-Aware Visual Cues for Scalable Annotation Review

arXiv:2605.12303v1 Announce Type: cross Abstract: High-quality labeled data is essential for training robust machine learning models, yet obtaining annotations at scale remains expensive. AI-assisted

From raw data to neutrino candidates: a neural-network pipeline for Baikal-GVD

arXiv:2605.11176v1 Announce Type: cross Abstract: We present a neural-network-based data processing pipeline for Baikal-GVD, designed to improve event reconstruction quality and accelerate neutrino ca

Fully AI-Generated Image Detection: Definition, Recent Advances and Challenges

arXiv:2502.19716v2 Announce Type: replace Abstract: Recent advances in visual generative models have enabled the creation of highly realistic, fully AI-generated images without relying on real source

FuTCR: Future-Targeted Contrast and Repulsion for Continual Panoptic Segmentation

arXiv:2605.12451v1 Announce Type: new Abstract: Continual Panoptic Segmentation (CPS) requires methods that can quickly adapt to new categories over time. The nature of this dense prediction task mean

G^2TR: Generation-Guided Visual Token Reduction for Separate-Encoder Unified Multimodal Models

arXiv:2605.12309v1 Announce Type: new Abstract: The development of separate-encoder Unified multimodal models (UMMs) comes with a rapidly growing inference cost due to dense visual token processing. I

GaitProtector: Impersonation-Driven Gait De-Identification via Training-Free Diffusion Latent Optimization

arXiv:2605.12431v1 Announce Type: new Abstract: Conventional gait de-identification methods often encounter an inherent trade-off: they either provide insufficient identity suppression or introduce sp

GATA2Floor: Graph attention for floor counting in street-view facades

arXiv:2605.11863v1 Announce Type: new Abstract: Automated analysis of building facades from street-level imagery has great potential for urban analytics, energy assessment, and emergency planning. How

GeoQuery: Geometry-Query Diffusion for Sparse-View Reconstruction

arXiv:2605.12399v1 Announce Type: new Abstract: 3D Gaussian Splatting (3DGS) has emerged as a prominent paradigm for 3D reconstruction and novel view synthesis. However, it remains vulnerable to sever

Gradient Clipping Beyond Vector Norms: A Spectral Approach for Matrix-Valued Parameters

arXiv:2605.11838v1 Announce Type: new Abstract: Gradient clipping is a standard safeguard for training neural networks under noisy, heavy-tailed stochastic gradients; yet, most clipping rules treat al

h-control: Training-Free Camera Control via Block-Conditional Gibbs Refinement

arXiv:2605.11871v1 Announce Type: new Abstract: Training-free camera control for pretrained flow-matching video generators is a partial-observation inverse problem: a depth-warped guidance video suppl

H2G: Hierarchy-Aware Hyperbolic Grouping for 3D Scenes

arXiv:2605.11967v1 Announce Type: new Abstract: Hierarchical 3D grouping aims to recover scene groups across multiple granularities, from fine object parts to complete objects, without relying on sema

HEPA: A Self-Supervised Horizon-Conditioned Event Predictive Architecture for Time Series

arXiv:2605.11130v1 Announce Type: new Abstract: Critical events in multivariate time series, from turbine failures to cardiac arrhythmias, demand accurate prediction, yet labeled data is scarce becaus

High-arity Sample Compression

arXiv:2605.12465v1 Announce Type: new Abstract: Recently, a series of works have started studying variations of concepts from learning theory for product spaces, which can be collected under the name

HorizonDrive: Self-Corrective Autoregressive World Model for Long-horizon Driving Simulation

arXiv:2605.11596v1 Announce Type: new Abstract: Closed-loop driving simulation requires real-time interaction beyond short offline clips, pushing current driving world models toward autoregressive (AR

Hyperbolic Concept Bottleneck Models

arXiv:2605.06440v2 Announce Type: replace-cross Abstract: Concept Bottleneck Models (CBMs) have become a popular approach to enable interpretability in neural networks by constraining classifier input

Hypernetworks for Dynamic Feature Selection

arXiv:2605.12278v1 Announce Type: new Abstract: Dynamic feature selection (DFS) is a machine learning framework in which features are acquired sequentially for individual samples under budget constrai

Improving the Performance and Learning Stability of Parallelizable RNNs Designed for Ultra-Low Power Applications

arXiv:2605.11855v1 Announce Type: new Abstract: Sequence learning is dominated by Transformers and parallelizable recurrent neural networks (RNNs) such as state-space models, yet learning long-term de

Information-Theoretic Generalization Bounds for Sequential Decision Making

arXiv:2605.12190v1 Announce Type: cross Abstract: Information-theoretic generalization bounds based on the supersample construction are a central tool for algorithm-dependent generalization analysis i

Information theoretic underpinning of self-supervised learning by clustering

arXiv:2605.11870v1 Announce Type: new Abstract: Self-supervised learning (SSL) is recognized as an essential tool for building foundation models for Artificial Intelligence applications. The advances

Integral Imprecise Probability Metrics

arXiv:2505.16156v3 Announce Type: replace-cross Abstract: Quantifying differences between probability distributions is fundamental to statistics and machine learning, primarily for comparing statistic

Interactive Mars Image Content-Based Search with Interpretable Machine Learning

arXiv:2402.16860v2 Announce Type: replace Abstract: The NASA Planetary Data System (PDS) hosts millions of images of planets, moons, and other bodies collected throughout many missions. The ever-expan

Interactive State Space Model with Cross-Modal Local Scanning for Depth Super-Resolution

arXiv:2605.11934v1 Announce Type: new Abstract: Guided depth super-resolution (GDSR) reconstructs HR depth maps from LR inputs with HR RGB guidance. Existing methods either model each modality indepen

Interesting paper. What I like about this is that it is a relatively low-commitment attention modification. I.e., one can use it during most…

researchsebastian-raschka--x

Interesting paper. What I like about this is that it is a relatively low-commitment attention modification. I.e., one can use it during most of training, switch back to vanilla attention near the end,

Interpretable EEG Microstate Discovery via Variational Deep Embedding: A Systematic Architecture Search with Multi-Quadrant Evaluation

arXiv:2605.10947v1 Announce Type: new Abstract: EEG microstate analysis segments continuous brain electrical activity into brief, quasi-stable topographic configurations that reflect discrete function

Interpretable Machine Learning for Spatial Science: A Lie-Algebraic Kernel for Rotationally Anisotropic Gaussian Processes

arXiv:2605.11179v1 Announce Type: cross Abstract: Many three-dimensional spatial fields are anisotropic, with directions of rapid and slow variation that need not align with the coordinate axes. Stand

Interpretable rainfall modelling reveals rapid reorganisation of Amazonian rainfall under vegetation loss

arXiv:2605.10948v1 Announce Type: cross Abstract: Understanding how vegetation loss alters rainfall remains a major challenge in climate and hydrological science, as deforestation modifies precipitati