AI研究論文每日精選

Seaweed-7B: Cost-Effective Training of Video Generation Foundation Model

GigaTok: Scaling Visual Tokenizers to 3 Billion Parameters for Autoregressive Image Generation

MineWorld: a Real-Time and Open-Source Interactive World Model on Minecraft

SQL-R1: Training Natural Language to SQL Reasoning Model By Reinforcement Learning

VLM-R1: A Stable and Generalizable R1-style Large Vision-Language Model

ZipIR: Latent Pyramid Diffusion Transformer for High-Resolution Image Restoration

PixelFlow: Pixel-Space Generative Models with Flow

Visual Chronicles: Using Multimodal LLMs to Analyze Massive Collections of Images

FlexIP: Dynamic Control of Preservation and Personality for Customized Image Generation

CoRAG: Collaborative Retrieval-Augmented Generation

ModernBERT or DeBERTaV3? Examining Architecture and Data Influence on Transformer Encoder Models Performance

Do PhD-level LLMs Truly Grasp Elementary Addition? Probing Rule Learning vs. Memorization in Large Language Models

In-2-4D: Inbetweening from Two Single-View Images to 4D Generation

Training-free Guidance in Text-to-Video Generation via Multimodal Planning and Structured Noise Initialization

UKBOB: One Billion MRI Labeled Masks for Generalizable 3D Medical Image Segmentation

BlenderGym: Benchmarking Foundational Model Systems for Graphics Editing

Latent Diffusion Autoencoders: Toward Efficient and Meaningful Unsupervised Representation Learning in Medical Imaging

SpecReason: Fast and Accurate Inference-Time Compute via Speculative Reasoning

SAEs Can Improve Unlearning: Dynamic Sparse Autoencoder Guardrails for Precision Unlearning in LLMs

InteractVLM: 3D Interaction Reasoning from 2D Foundational Models