AI 연구 논문 데일리

Movie Gen: A Cast of Media Foundation Models

MixEval-X: Any-to-Any Evaluations from Real-World Data Mixtures

JudgeBench: A Benchmark for Evaluating LLM-based Judges

Fluid: Scaling Autoregressive Text-to-image Generative Models with Continuous Tokens

Roadmap towards Superhuman Speech Understanding using Large Language Models

MobA: A Two-Level Agent System for Efficient Mobile Task Automation

WorldCuisines: A Massive-Scale Benchmark for Multilingual and Multicultural Visual Question Answering on Global Cuisines

Harnessing Webpage UIs for Text-Rich Visual Understanding

Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation

DreamVideo-2: Zero-Shot Subject-Driven Video Customization with Precise Motion Control

MoH: Multi-Head Attention as Mixture-of-Head Attention

MMed-RAG: Versatile Multimodal RAG System for Medical Vision Language Models

BenTo: Benchmark Task Reduction with In-Context Transferability

PopAlign: Diversifying Contrasting Patterns for a More Comprehensive Alignment

A Comparative Study on Reasoning Patterns of OpenAI's o1 Model

A Unified View of Delta Parameter Editing in Post-Trained Large-Scale Models

Do LLMs Have Political Correctness? Analyzing Ethical Biases and Jailbreak Vulnerabilities in AI Systems

VidPanos: Generative Panoramic Videos from Casual Panning Videos

FlatQuant: Flatness Matters for LLM Quantization

Failing Forward: Improving Generative Error Correction for ASR with Synthetic Data and Retrieval Augmentation

Can MLLMs Understand the Deep Implication Behind Chinese Images?

Remember, Retrieve and Generate: Understanding Infinite Visual Concepts as Your Personalized Assistant

MedMobile: A mobile-sized language model with expert-level clinical capabilities

Retrospective Learning from Interactions

γ-MoD: Exploring Mixture-of-Depth Adaptation for Multimodal Large Language Models

MuVi: Video-to-Music Generation with Semantic Alignment and Rhythmic Synchronization

Open Materials 2024 (OMat24) Inorganic Materials Dataset and Models

LoLDU: Low-Rank Adaptation via Lower-Diag-Upper Decomposition for Parameter-Efficient Fine-Tuning

Long-LRM: Long-sequence Large Reconstruction Model for Wide-coverage Gaussian Splats

Toward Guidance-Free AR Visual Generation via Condition Contrastive Alignment

AERO: Softmax-Only LLMs for Efficient Private Inference

Minimum Tuning to Unlock Long Output from LLMs with High Quality Data as the Key

TransAgent: Transfer Vision-Language Foundation Models with Heterogeneous Agent Collaboration

SBI-RAG: Enhancing Math Word Problem Solving for Students through Schema-Based Instruction and Retrieval-Augmented Generation