AI 연구 논문 데일리

VidEgoThink: Assessing Egocentric Video Understanding Capabilities for Embodied AI

HumanEval-V: Evaluating Visual Understanding and Reasoning Abilities of Large Multimodal Models Through Coding Tasks

The Curse of Multi-Modalities: Evaluating Hallucinations of Large Multimodal Models across Language, Visual, and Audio

DocLayout-YOLO: Enhancing Document Layout Analysis through Diverse Synthetic Data and Global-to-Local Adaptive Perception

Revealing the Barriers of Language Agents in Planning

Exploring Model Kinship for Merging Large Language Models

Simplifying, Stabilizing and Scaling Continuous-Time Consistency Models

Large Language Model Evaluation via Matrix Nuclear-Norm

Improving Long-Text Alignment for Text-to-Image Diffusion Models

ProSA: Assessing and Understanding the Prompt Sensitivity of LLMs

Controllable Safety Alignment: Inference-Time Adaptation to Diverse Safety Requirements

DyVo: Dynamic Vocabularies for Learned Sparse Retrieval with Entities

ZipVL: Efficient Large Vision-Language Models with Dynamic Token Sparsification and KV Cache Compression

Stabilize the Latent Space for Image Autoregressive Modeling: A Unified Perspective

Neural Metamorphosis

ChroKnowledge: Unveiling Chronological Knowledge of Language Models in Multiple Domains

Tracking Universal Features Through Fine-Tuning and Model Merging

WorldMedQA-V: a multilingual, multimodal medical examination dataset for multimodal language models evaluation

Insights from the Inverse: Reconstructing LLM Training Goals Through Inverse RL

OMCAT: Omni Context Aware Transformer

FLARE: Faithful Logic-Aided Reasoning and Exploration

Taming Overconfidence in LLMs: Reward Calibration in RLHF

From Commands to Prompts: LLM-based Semantic File System for AIOS