KI-Forschungspapiere Täglich

SymDPO: Boosting In-Context Learning of Large Multimodal Models with Symbol Demonstration Direct Preference Optimization

SageAttention2 Technical Report: Accurate 4 Bit Attention for Plug-and-play Inference Acceleration

VBench++: Comprehensive and Versatile Benchmark Suite for Video Generative Models

VideoAutoArena: An Automated Arena for Evaluating Large Multimodal Models in Video Analysis through User Simulation

When Precision Meets Position: BFloat16 Breaks Down RoPE in Long-Context Training

SAMURAI: Adapting Segment Anything Model for Zero-Shot Visual Tracking with Motion-Aware Memory

Is Your LLM Secretly a World Model of the Internet? Model-Based Planning for Web Agents

Stylecodes: Encoding Stylistic Information For Image Generation

ViBe: A Text-to-Video Benchmark for Evaluating Hallucination in Large Multimodal Models

Loss-to-Loss Prediction: Scaling Laws for All Datasets

Generating Compositional Scenes via Text-to-image RGBA Instance Generation

ORID: Organ-Regional Information Driven Framework for Radiology Report Generation