AI研究論文每日精選

Phi-4 Technical Report

InternLM-XComposer2.5-OmniLive: A Comprehensive Multimodal System for Long-term Streaming Video and Audio Interactions

Euclid: Supercharging Multimodal LLMs with Synthetic High-Fidelity Visual Descriptions

Lyra: An Efficient and Speech-Centric Framework for Omni-Cognition

Multimodal Latent Language Modeling with Next-Token Diffusion

AgentTrek: Agent Trajectory Synthesis via Guiding Replay with Web Tutorials

SnapGen: Taming High-Resolution Text-to-Image Models for Mobile Devices with Efficient Architectures and Training

EasyRef: Omni-Generalized Group Image Reference for Diffusion Models via Multimodal LLM

JuStRank: Benchmarking LLM Judges for System Ranking

PIG: Physics-Informed Gaussians as Adaptive Parametric Mesh Representations

Neural LightRig: Unlocking Accurate Object Normal and Material Estimation with Multi-Light Diffusion

VisionArena: 230K Real World User-VLM Conversations with Preference Labels

Learned Compression for Compressed Learning

Arbitrary-steps Image Super-resolution via Diffusion Inversion

OLA-VLM: Elevating Visual Perception in Multimodal LLMs with Auxiliary Embedding Distillation

RuleArena: A Benchmark for Rule-Guided Reasoning with LLMs in Real-World Scenarios

Normalizing Flows are Capable Generative Models

Word Sense Linking: Disambiguating Outside the Sandbox

DisPose: Disentangling Pose Guidance for Controllable Human Image Animation

LoRACLR: Contrastive Adaptation for Customization of Diffusion Models

FreeSplatter: Pose-free Gaussian Splatting for Sparse-view 3D Reconstruction

The Impact of Copyrighted Material on Large Language Models: A Norwegian Perspective

ONEBench to Test Them All: Sample-Level Benchmarking Over Open-Ended Capabilities

Gaze-LLE: Gaze Target Estimation via Large-Scale Learned Encoders

SAME: Learning Generic Language-Guided Visual Navigation with State-Adaptive Mixture of Experts

Shiksha: A Technical Domain focused Translation Dataset and Model for Indian Languages