Knowledge catalogue
arXiv:2605.09449v1 Announce Type: new Abstract: Recent multimodal large language models (MLLMs) have made remarkable progress in visual understanding and language-based reasoning, yet they lack a pers
arXiv:2601.11042v2 Announce Type: replace-cross Abstract: Sequential knowledge editing in large language models often causes catastrophic collapse of the model's general abilities, especially for para
arXiv:2508.08441v3 Announce Type: replace-cross Abstract: Automated molecular structure elucidation remains challenging, as existing approaches often depend on pre-compiled databases or restrict thems
arXiv:2502.18334v5 Announce Type: replace Abstract: Graph-based learning excels at capturing interaction patterns in diverse domains like recommendation, fraud detection, and particle physics. However
arXiv:2605.08611v1 Announce Type: new Abstract: Current language model memory systems store what happened but not how it felt. This distinction -- between semantic memory (knowing about a past event)
arXiv:2604.02474v2 Announce Type: replace Abstract: Dynamical systems describe how a physical system evolves over time. Physical processes can evolve faster or slower in different environmental condit
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
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
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 (
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
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-
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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.
arXiv:2602.17251v2 Announce Type: replace Abstract: Electroencephalography (EEG) foundation models (EFMs) have shown strong potential for transferable representation learning, yet their adaptation in
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
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
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
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
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
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)
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
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
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
Yann LeCun closed 1.03B for AMI Labs on March 10. Three days later, this paper dropped from his NYU collaborators. 15M parameters. Single GPU. A few hours of training. LeWorldModel is the first JEPA t
I promise this will be the best 20 min you spend today! Robotics: Endgame, the sequel to my last year's Sequoia AI Ascent talk, 'Physical Turing Test'. I laid out the roadmap for solving Physical AGI
arXiv:2605.04916v1 Announce Type: cross Abstract: Inductive Logic Programming (ILP) learns interpretable logical rules from data. Existing methods are transductive: their learned parameters are bound
arXiv:2601.20634v2 Announce Type: replace Abstract: Virtual sensors replace expensive physical sensors in critical applications through machine learning by predicting target signals from available mea
arXiv:2605.05123v1 Announce Type: new Abstract: In offline-to-online reinforcement learning (O2O-RL), policies are first safely trained offline using previously collected datasets and then further fin
arXiv:2605.04911v1 Announce Type: new Abstract: Tabular data synthesis aims to generate high-quality data while preserving privacy. However, we find that existing tabular generative models exhibit a c
arXiv:2605.05012v1 Announce Type: new Abstract: Texture classification is a pivotal task in computer vision, presenting unique challenges due to high inter-class similarity and the sensitivity of stru