research

Easy Samples Are All You Need: Self-Evolving LLMs via Data-Efficient Reinforcement Learning

arXiv:2604.18639v1 Announce Type: cross Abstract: Previous LLMs-based RL studies typically follow either supervised learning with high annotation costs, or unsupervised paradigms using voting or entro

Research

Efficient Autoregressive Inference for Transformer Probabilistic Models

DGX agent

arXiv:2510.09477v2 Announce Type: replace-cross Abstract: Set-based transformer models for amortized probabilistic inference and meta-learning, such as neural processes, prior-fitted networks, and tab

researcharxiv-cs-lg

22 Apr 2026

EgoMotion: Hierarchical Reasoning and Diffusion for Egocentric Vision-Language Motion Generation

arXiv:2604.19105v1 Announce Type: new Abstract: Faithfully modeling human behavior in dynamic environments is a foundational challenge for embodied intelligence. While conditional motion synthesis has

EgoSelf: From Memory to Personalized Egocentric Assistant

arXiv:2604.19564v1 Announce Type: cross Abstract: Egocentric assistants often rely on first-person view data to capture user behavior and context for personalized services. Since different users exhib

EHRAG: Bridging Semantic Gaps in Lightweight GraphRAG via Hybrid Hypergraph Construction and Retrieval

arXiv:2604.17458v2 Announce Type: replace Abstract: Graph-based Retrieval-Augmented Generation (GraphRAG) enhances LLMs by structuring corpus into graphs to facilitate multi-hop reasoning. While recen

Enforcing Reciprocity in Operator Learning for Seismic Wave Propagation

arXiv:2602.11631v2 Announce Type: replace-cross Abstract: Accurate and efficient wavefield modeling underpins seismic structure and source studies. Traditional methods comply with physical laws but ar

Enhancing Unsupervised Keyword Extraction in Academic Papers through Integrating Highlights with Abstract

arXiv:2604.19505v1 Announce Type: cross Abstract: Automatic keyword extraction from academic papers is a key area of interest in natural language processing and information retrieval. Although previou

Epistemic orientation in parliamentary discourse is associated with deliberative democracy

arXiv:2604.19699v1 Announce Type: new Abstract: The pursuit of truth is central to democratic deliberation and governance, yet political discourse reflects varying epistemic orientations, ranging from

Epistemic Skills: Reasoning about Knowledge and Oblivion

arXiv:2504.01733v4 Announce Type: replace Abstract: This paper presents a class of epistemic logics that captures the dynamics of acquiring knowledge and descending into oblivion, while incorporating

EPS: Efficient Patch Sampling for Video Overfitting in Deep Super-Resolution Model Training

arXiv:2411.16312v2 Announce Type: replace Abstract: Leveraging the overfitting property of deep neural networks (DNNs) is trending in video delivery systems to enhance video quality within bandwidth l

Error-free Training for MedMNIST Datasets

arXiv:2604.18916v1 Announce Type: new Abstract: In this paper, we introduce a new concept called Artificial Special Intelligence by which Machine Learning models for the classification problem can be

EssayCBM: Rubric-Aligned Concept Bottleneck Models for Transparent Essay Grading

arXiv:2512.20817v2 Announce Type: replace Abstract: Automated essay scoring (AES) has advanced significantly with neural language models, yet most systems remain opaque, offering little visibility int

Evaluating LLM-Generated Obfuscated XSS Payloads for Machine Learning-Based Detection

arXiv:2604.19526v1 Announce Type: cross Abstract: Cross-site scripting (XSS) remains a persistent web security vulnerability, especially because obfuscation can change the surface form of a malicious

Evaluation of Winning Solutions of 2025 Low Power Computer Vision Challenge

arXiv:2604.19054v1 Announce Type: new Abstract: The IEEE Low-Power Computer Vision Challenge (LPCVC) aims to promote the development of efficient vision models for edge devices, balancing accuracy wit

Experiments or Outcomes? Probing Scientific Feasibility in Large Language Models

arXiv:2604.18786v1 Announce Type: cross Abstract: Scientific feasibility assessment asks whether a claim is consistent with established knowledge and whether experimental evidence could support or ref

Face Anything: 4D Face Reconstruction from Any Image Sequence

arXiv:2604.19702v1 Announce Type: new Abstract: Accurate reconstruction and tracking of dynamic human faces from image sequences is challenging because non-rigid deformations, expression changes, and

Fast estimation of Gaussian mixture components via centering and singular value thresholding

arXiv:2604.19091v1 Announce Type: cross Abstract: Estimating the number of components is a fundamental challenge in unsupervised learning, particularly when dealing with high-dimensional data with man

Feature Perturbation Pool-based Fusion Network for Unified Multi-Class Industrial Defect Detection

arXiv:2604.19259v1 Announce Type: new Abstract: Multi-class defect detection constitutes a critical yet challenging task in industrial quality inspection, where existing approaches typically suffer fr

FedSEA: Achieving Benefit of Parallelization in Federated Online Learning

arXiv:2604.19336v1 Announce Type: new Abstract: Online federated learning (OFL) has emerged as a popular framework for decentralized decision-making over continuous data streams without compromising c

Fine-Tuning Code Language Models to Detect Cross-Language Bugs

arXiv:2507.21954v2 Announce Type: replace-cross Abstract: Multilingual programming, which involves using multiple programming languages (PLs) in a single project, is increasingly common due to its ben

Finite-dimensional approximations of push-forwards on locally analytic functionals

arXiv:2404.10769v4 Announce Type: replace-cross Abstract: This paper develops a functional-analytic framework for approximating the push-forward induced by an analytic map from finitely many samples.

Forward Dynamics of Variable Topology Mechanisms - The Case of Constraint Activation

arXiv:2604.19419v1 Announce Type: new Abstract: Many mechanical systems exhibit changes in their kinematic topology altering the mobility. Ideal contact is the best known cause, but also stiction and

researcharxiv-cs-ro

GaiaFlow: Semantic-Guided Diffusion Tuning for Carbon-Frugal Search

arXiv:2602.15423v3 Announce Type: replace-cross Abstract: As the burgeoning power requirements of sophisticated neural architectures escalate, the information retrieval community has recognized ecolog

GAIN: Multiplicative Modulation for Domain Adaptation

arXiv:2604.04516v2 Announce Type: replace-cross Abstract: Adapting LLMs to new domains causes forgetting because standard methods (e.g., full fine-tuning, LoRA) inject new directions into the weight s

GAIR: Location-Aware Self-Supervised Contrastive Pre-Training with Geo-Aligned Implicit Representations

arXiv:2503.16683v2 Announce Type: replace-cross Abstract: Vision Transformer (ViT) has been widely used in computer vision tasks with excellent results by providing representations for a whole image o

Generalization at the Edge of Stability

arXiv:2604.19740v1 Announce Type: cross Abstract: Training modern neural networks often relies on large learning rates, operating at the edge of stability, where the optimization dynamics exhibit osci

Generative Texture Filtering

arXiv:2604.19039v1 Announce Type: new Abstract: We present a generative method for texture filtering, which exhibits surprisingly good performance and generalizability. Our core idea is to empower tex

Geometric Decoupling: Diagnosing the Structural Instability of Latent

arXiv:2604.18804v1 Announce Type: cross Abstract: Latent Diffusion Models (LDMs) achieve high-fidelity synthesis but suffer from latent space brittleness, causing discontinuous semantic jumps during e

Geometry-Guided Self-Supervision for Ultra-Fine-Grained Recognition with Limited Data

arXiv:2604.19345v1 Announce Type: new Abstract: This paper investigates the intrinsic geometrical features of highly similar objects and introduces a general self-supervised framework called the Geome

Global Optimization of Gaussian Process Acquisition Functions Using a Piecewise-Linear Kernel Approximation

arXiv:2410.16893v3 Announce Type: replace-cross Abstract: Bayesian optimization relies on iteratively constructing and optimizing an acquisition function. The latter turns out to be a challenging, non

GOLD-BEV: GrOund and aeriaL Data for Dense Semantic BEV Mapping of Dynamic Scenes

arXiv:2604.19411v1 Announce Type: cross Abstract: Understanding road scenes in a geometrically consistent, scene-centric representation is crucial for planning and mapping. We present GOLD-BEV, a fram

Headlines You Won't Forget: Can Pronoun Insertion Increase Memorability?

arXiv:2604.19189v1 Announce Type: new Abstract: For news headlines to influence beliefs and drive action, relevant information needs to be retained and retrievable from memory. In this probing study w

Highly Efficient and Effective LLMs with Multi-Boolean Architectures

arXiv:2505.22811v5 Announce Type: replace-cross Abstract: Weight binarization has emerged as a promising strategy to reduce the complexity of large language models (LLMs). Existing approaches fall int

https://sakana.ai/careers/#principal-platform-engineer

Sakana AI is hiring for a Principal Platform Engineer position, as shared by David Ha on X. This role likely involves building and maintaining infrastructure, systems, and tools that support the compa

researchdavid-ha--x

Hybrid Architectures for Language Models: Systematic Analysis and Design Insights

arXiv:2510.04800v3 Announce Type: replace Abstract: Recent progress in large language models demonstrates that hybrid architectures--combining self-attention mechanisms with structured state space mod

I am looking for a Principal Platform Engineer to join us in Tokyo and build the core data infrastructure for Japan’s most critical systems.…

I am looking for a Principal Platform Engineer to join us in Tokyo and build the core data infrastructure for Japan’s most critical systems. Read the full mission and apply here: https://sakana.ai/car

researchdavid-ha--x

Idle is the New Sleep: Configuration-Aware Alternative to Powering Off FPGA-Based DL Accelerators During Inactivity

arXiv:2407.12027v2 Announce Type: replace-cross Abstract: In the rapidly evolving Internet of Things (IoT) domain, we concentrate on enhancing energy efficiency in Deep Learning accelerators on FPGA-b

Imagine every pixel on your screen, streamed live directly from a model. No HTML, no layout engine, no code. Just exactly what you want to s…

Imagine every pixel on your screen, streamed live directly from a model. No HTML, no layout engine, no code. Just exactly what you want to see. @eddiejiao_obj, @drewocarr and I built a prototype to se

researchkarpathy--x

Impact of large language models on peer review opinions from a fine-grained perspective: Evidence from top conference proceedings in AI

arXiv:2604.19578v1 Announce Type: cross Abstract: With the rapid advancement of Large Language Models (LLMs), the academic community has faced unprecedented disruptions, particularly in the realm of a

Improved Anomaly Detection in Medical Images via Mean Shift Density Enhancement

arXiv:2604.19191v1 Announce Type: cross Abstract: Anomaly detection in medical imaging is essential for identifying rare pathological conditions, particularly when annotated abnormal samples are limit

Improvements to the post-processing of weather forecasts using machine learning and feature selection

arXiv:2604.19340v1 Announce Type: cross Abstract: This study aims to develop and improve machine learning-based post-processing models for precipitation, temperature, and wind speed predictions using

“In the end the Party would announce that two and two made five, and you would have to believe it.” -George Orwell, 1984.

“In the end the Party would announce that two and two made five, and you would have to believe it.” -George Orwell, 1984. RFK Jr: 'President Trump has a different way of calculating percentages. If yo

researchyann-lecun--x

INT3 compression+fused metal kernels [R]

researchr-machinelearning

INT3 compression with fused Metal kernels enables large language models to compute attention operations directly on compressed (INT3/INT4) key-value cache representations using custom GPU kernels that

Investigating the structure of emotions by analyzing similarity and association of emotion words

arXiv:2602.06430v2 Announce Type: replace-cross Abstract: In the field of natural language processing, some studies have attempted sentiment analysis on text by handling emotions as explanatory or res

IonMorphNet: Generalizable Learning of Ion Image Morphologies for Peak Picking in Mass Spectrometry Imaging

arXiv:2604.19369v1 Announce Type: new Abstract: Peak picking is a fundamental preprocessing step in Mass Spectrometry Imaging (MSI), where each sample is represented by hundreds to thousands of ion im

KG-ViP: Bridging Knowledge Grounding and Visual Perception in Multi-modal LLMs for Visual Question Answering

arXiv:2601.11632v2 Announce Type: replace Abstract: Multi-modal Large Language Models (MLLMs) for Visual Question Answering (VQA) often suffer from dual limitations: knowledge hallucination and insuff

Learning Evolution via Optimization Knowledge Adaptation

arXiv:2501.02200v2 Announce Type: replace-cross Abstract: The iterative search process of evolutionary algorithms (EAs) encapsulates optimization knowledge within historical populations and fitness ev

Los Angeles is finally going underground

Los Angeles deserves its reputation as the quintessential car city—the rhythms of its 2,200 square miles are dictated by wide boulevards and concrete arcs of freeways. But it once had a world-class ra

researchmit-tech-review