Search: “pretraining”

Curing Miracle Steps in LLM Mathematical Reasoning with Rubric Rewards

arXiv:2510.07774v3 Announce Type: replace Abstract: In this paper, we observe that current models are susceptible to reward hacking, leading to a substantial overestimation of a model's reasoning abil

researcharxiv-cs-cl

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Differential privacy representation geometry for medical image analysis

DGX agent

arXiv:2603.01098v2 Announce Type: replace-cross Abstract: Differential privacy (DP)'s effect in medical imaging is typically evaluated only through end-to-end performance, leaving the mechanism of pri

researcharxiv-cs-ai

20 Apr 2026

Do LLMs Really Know What They Don't Know? Internal States Mainly Reflect Knowledge Recall Rather Than Truthfulness

arXiv:2510.09033v3 Announce Type: replace Abstract: Recent work suggests that LLMs 'know what they don't know', positing that hallucinated and factually correct outputs arise from distinct internal pr

researcharxiv-cs-cl

Find, Fix, Reason: Context Repair for Video Reasoning

arXiv:2604.16243v1 Announce Type: new Abstract: Reinforcement learning has advanced video reasoning in large multi-modal models, yet dominant pipelines either rely on on-policy self-exploration, which

safetyarxiv-cs-cv

GenHSI: Controllable Generation of Human-Scene Interaction Videos

arXiv:2506.19840v2 Announce Type: replace Abstract: Large-scale pre-trained video diffusion models have exhibited remarkable capabilities in diverse video generation. However, existing solutions face

MMGait: Towards Multi-Modal Gait Recognition

arXiv:2604.15979v1 Announce Type: new Abstract: Gait recognition has emerged as a powerful biometric technique for identifying individuals at a distance without requiring user cooperation. Most existi

Neural Continuous-Time Markov Chain: Discrete Diffusion via Decoupled Jump Timing and Direction

arXiv:2604.15694v1 Announce Type: new Abstract: Discrete diffusion models based on continuous-time Markov chains (CTMCs) have shown strong performance on language and discrete data generation, yet exi

Tutorials

OT on the Map: Quantifying Domain Shifts in Geographic Space

arXiv:2604.16220v1 Announce Type: new Abstract: In computer vision and machine learning for geographic data, out-of-domain generalization is a pervasive challenge, arising from uneven global data cove

tutorialsarxiv-cs-lg

Scaling Behaviors of LLM Reinforcement Learning Post-Training: An Empirical Study in Mathematical Reasoning

arXiv:2509.25300v4 Announce Type: replace-cross Abstract: While scaling laws for large language models (LLMs) during pre-training have been extensively studied, their behavior under reinforcement lear

researcharxiv-cs-ai

Applications

Sketching the Readout of Large Language Models for Scalable Data Attribution and Valuation

arXiv:2604.16197v1 Announce Type: new Abstract: Data attribution and valuation are critical for understanding data-model synergy for Large Language Models (LLMs), yet existing gradient-based methods s

applicationsarxiv-cs-lg

SurgMotion: A Video-Native Foundation Model for Universal Understanding of Surgical Videos

arXiv:2602.05638v3 Announce Type: replace Abstract: While foundation models have advanced surgical video analysis, current approaches rely predominantly on pixel-level reconstruction objectives that w

UA-Net: Uncertainty-Aware Network for TRISO Image Semantic Segmentation

arXiv:2604.15542v1 Announce Type: new Abstract: Tristructural isotropic (TRISO)-coated particle fuels undergo dimensional changes and chemical reactions during high-temperature neutron irradiation. Po

Winner of CVPR2026 NTIRE Challenge on Image Shadow Removal: Semantic and Geometric Guidance for Shadow Removal via Cascaded Refinement

arXiv:2604.16177v1 Announce Type: new Abstract: We present a three-stage progressive shadow-removal pipeline for the CVPR2026 NTIRE WSRD+ challenge. Built on OmniSR, our method treats deshadowing as i

Applications

Beyond Independent Frames: Latent Attention Masked Autoencoders for Multi-View Echocardiography

arXiv:2604.15096v1 Announce Type: new Abstract: Echocardiography is a widely used modality for cardiac assessment due to its non-invasive and cost-effective nature, but the sparse and heterogeneous sp

applicationsarxiv-cs-cv

Beyond Prompts: Unconditional 3D Inversion for Out-of-Distribution Shapes

arXiv:2604.14914v1 Announce Type: new Abstract: Text-driven inversion of generative models is a core paradigm for manipulating 2D or 3D content, unlocking numerous applications such as text-based edit

Class Unlearning via Depth-Aware Removal of Forget-Specific Directions

arXiv:2604.15166v1 Announce Type: new Abstract: Machine unlearning aims to remove targeted knowledge from a trained model without the cost of retraining from scratch. In class unlearning, however, red

Co-distilled attention guided masked image modeling with noisy teacher for self-supervised learning on medical images

arXiv:2604.14506v1 Announce Type: new Abstract: Masked image modeling (MIM) is a highly effective self-supervised learning (SSL) approach to extract useful feature representations from unannotated dat

CobwebTM: Probabilistic Concept Formation for Lifelong and Hierarchical Topic Modeling

arXiv:2604.14489v1 Announce Type: new Abstract: Topic modeling seeks to uncover latent semantic structure in text corpora with minimal supervision. Neural approaches achieve strong performance but req

Agents

Empowerment Gain and Causal Model Construction: Children and adults are sensitive to controllability and variability in their causal interventions

arXiv:2512.08230v2 Announce Type: replace Abstract: Learning about the causal structure of the world is a fundamental problem for human cognition. Causal models and especially causal learning have pro

agentsarxiv-cs-ai

Enhancing LLM-based Search Agents via Contribution Weighted Group Relative Policy Optimization

arXiv:2604.14267v1 Announce Type: new Abstract: Search agents extend Large Language Models (LLMs) beyond static parametric knowledge by enabling access to up-to-date and long-tail information unavaila

safetyarxiv-cs-lg

Local Ai

GlobalSplat: Efficient Feed-Forward 3D Gaussian Splatting via Global Scene Tokens

arXiv:2604.15284v1 Announce Type: new Abstract: The efficient spatial allocation of primitives serves as the foundation of 3D Gaussian Splatting, as it directly dictates the synergy between representa

local-aiarxiv-cs-cv

KVNN: Learnable Multi-Kernel Volterra Neural Networks

arXiv:2604.15141v1 Announce Type: new Abstract: Higher-order learning is fundamentally rooted in exploiting compositional features. It clearly hinges on enriching the representation by more elaborate

Layered Mutability: Continuity and Governance in Persistent Self-Modifying Agents

arXiv:2604.14717v1 Announce Type: cross Abstract: Persistent language-model agents increasingly combine tool use, tiered memory, reflective prompting, and runtime adaptation. In such systems, behavior

safetyarxiv-cs-lg

Portfolio Optimization Proxies under Label Scarcity and Regime Shifts via Bayesian and Deterministic Students under Semi-Supervised Sandwich Training

arXiv:2604.14206v1 Announce Type: new Abstract: This paper proposes a machine learning assisted portfolio optimization framework designed for low data environments and regime uncertainty. We construct

Predicting Post-Traumatic Epilepsy from Clinical Records using Large Language Model Embeddings

arXiv:2604.14547v1 Announce Type: new Abstract: Objective: Post-traumatic epilepsy (PTE) is a debilitating neurological disorder that develops after traumatic brain injury (TBI). Early prediction of P

Tutorials

ReSS: Learning Reasoning Models for Tabular Data Prediction via Symbolic Scaffold

arXiv:2604.13392v1 Announce Type: new Abstract: Tabular data remains prevalent in high-stakes domains such as healthcare and finance, where predictive models are expected to provide both high accuracy

tutorialsarxiv-cs-ai

Retrieve, Then Classify: Corpus-Grounded Automation of Clinical Value Set Authoring

arXiv:2604.14616v1 Announce Type: new Abstract: Clinical value set authoring -- the task of identifying all codes in a standardized vocabulary that define a clinical concept -- is a recurring bottlene

Robustness of Vision Foundation Models to Common Perturbations

arXiv:2604.14973v1 Announce Type: cross Abstract: A vision foundation model outputs an embedding vector for an image, which can be affected by common editing operations (e.g., JPEG compression, bright

SegWithU: Uncertainty as Perturbation Energy for Single-Forward-Pass Risk-Aware Medical Image Segmentation

arXiv:2604.15271v1 Announce Type: new Abstract: Reliable uncertainty estimation is critical for medical image segmentation, where automated contours feed downstream quantification and clinical decisio

Shuffle the Context: RoPE-Perturbed Self-Distillation for Long-Context Adaptation

arXiv:2604.14339v1 Announce Type: new Abstract: Large language models (LLMs) increasingly operate in settings that require reliable long-context understanding, such as retrieval-augmented generation a

TokenLight: Precise Lighting Control in Images using Attribute Tokens

arXiv:2604.15310v1 Announce Type: new Abstract: This paper presents a method for image relighting that enables precise and continuous control over multiple illumination attributes in a photograph. We

When Does Content-Based Routing Work? Representation Requirements for Selective Attention in Hybrid Sequence Models

arXiv:2603.20997v2 Announce Type: replace Abstract: We identify a routing paradox in hybrid sequence models: content-based routing - deciding which tokens deserve expensive attention - requires pairwi

A1: A Fully Transparent Open-Source, Adaptive and Efficient Truncated Vision-Language-Action Model

arXiv:2604.05672v3 Announce Type: replace Abstract: Vision-Language-Action (VLA) models have emerged as a powerful paradigm for open-world robot manipulation, but their practical deployment is often c

researcharxiv-cs-ro

Action Images: End-to-End Policy Learning via Multiview Video Generation

arXiv:2604.06168v2 Announce Type: replace Abstract: World action models (WAMs) have emerged as a promising direction for robot policy learning, as they can leverage powerful video backbones to model t

safetyarxiv-cs-cv

Concrete Jungle: Towards Concreteness Paved Contrastive Negative Mining for Compositional Understanding

arXiv:2604.13313v1 Announce Type: new Abstract: Vision-Language Models demonstrate remarkable capabilities but often struggle with compositional reasoning, exhibiting vulnerabilities regarding word or

Cyclic 2.5D Perceptual Loss for Cross-Modal 3D Medical Image Synthesis: T1w MRI to Tau PET

arXiv:2406.12632v3 Announce Type: replace-cross Abstract: Positron emission tomography (PET) provides molecular biomarkers for Alzheimer's disease and related dementias (ADRD) and is increasingly used

Deep Spatially-Regularized and Superpixel-Based Diffusion Learning for Unsupervised Hyperspectral Image Clustering

arXiv:2604.13307v1 Announce Type: new Abstract: An unsupervised framework for hyperspectral image (HSI) clustering is proposed that incorporates masked deep representation learning with diffusion-base

DiffMagicFace: Identity Consistent Facial Editing of Real Videos

arXiv:2604.13841v1 Announce Type: new Abstract: Text-conditioned image editing has greatly benefitted from the advancements in Image Diffusion Models. However, extending these techniques to facial vid

F-Actor: Controllable Conversational Behaviour in Full-Duplex Models

arXiv:2601.11329v3 Announce Type: replace Abstract: Spoken conversational systems require more than accurate speech generation to have human-like conversations: to feel natural and engaging, they must

fMRI-LM: Towards a Universal Foundation Model for Language-Aligned fMRI Understanding

arXiv:2511.21760v3 Announce Type: replace Abstract: Recent advances in multimodal large language models (LLMs) have enabled unified reasoning across images, audio, and video, but extending such capabi

GeoBridge: A Semantic-Anchored Multi-View Foundation Model Bridging Images and Text for Geo-Localization

arXiv:2512.02697v3 Announce Type: replace Abstract: Cross-view geo-localization infers a location by retrieving geo-tagged reference images that visually correspond to a query image. However, the trad

Jump-Start Reinforcement Learning with Vision-Language-Action Regularization

arXiv:2604.13733v1 Announce Type: new Abstract: Reinforcement learning (RL) enables high-frequency, closed-loop control for robotic manipulation, but scaling to long-horizon tasks with sparse or imper

safetyarxiv-cs-lg

Tutorials

Multitasking Embedding for Embryo Blastocyst Grading Prediction (MEmEBG)

arXiv:2604.13217v1 Announce Type: new Abstract: Reliable evaluation of blastocyst quality is critical for the success of in vitro fertilization (IVF) treatments. Current embryo grading practices prima

tutorialsarxiv-cs-cv

Native Hybrid Attention for Efficient Sequence Modeling

arXiv:2510.07019v3 Announce Type: replace Abstract: Transformers excel at sequence modeling but face quadratic complexity, while linear attention offers improved efficiency but often compromises recal

researcharxiv-cs-cl

Local Ai

Simplicity Prevails: The Emergence of Generalizable AIGI Detection in Visual Foundation Models

arXiv:2602.01738v2 Announce Type: replace Abstract: While specialized detectors for AI-Generated Images (AIGI) achieve near-perfect accuracy on curated benchmarks, they suffer from a dramatic performa

local-aiarxiv-cs-cv

Training-Free Semantic Multi-Object Tracking with Vision-Language Models

arXiv:2604.14074v1 Announce Type: new Abstract: Semantic Multi-Object Tracking (SMOT) extends multi-object tracking with semantic outputs such as video summaries, instance-level captions, and interact

UNRIO: Uncertainty-Aware Velocity Learning for Radar-Inertial Odometry

model-releasesarxiv-cs-ro

arXiv:2604.13584v1 Announce Type: new Abstract: We present UNRIO, an uncertainty-aware radar-inertial odometry system that estimates ego-velocity directly from raw mmWave radar IQ signals rather than

VGGT-Segmentor: Geometry-Enhanced Cross-View Segmentation

arXiv:2604.13596v1 Announce Type: new Abstract: Instance-level object segmentation across disparate egocentric and exocentric views is a fundamental challenge in visual understanding, critical for app