KI-Forschungspapiere Täglich

Block Transformer: Global-to-Local Language Modeling for Fast Inference

Parrot: Multilingual Visual Instruction Tuning

Mobile-Agent-v2: Mobile Device Operation Assistant with Effective Navigation via Multi-Agent Collaboration

Ouroboros3D: Image-to-3D Generation via 3D-aware Recursive Diffusion

Audio Mamba: Bidirectional State Space Model for Audio Representation Learning

PosterLLaVa: Constructing a Unified Multi-modal Layout Generator with LLM

LiveSpeech: Low-Latency Zero-shot Text-to-Speech via Autoregressive Modeling of Audio Discrete Codes

Searching Priors Makes Text-to-Video Synthesis Better

Scaling Laws for Reward Model Overoptimization in Direct Alignment Algorithms

Item-Language Model for Conversational Recommendation

PLaD: Preference-based Large Language Model Distillation with Pseudo-Preference Pairs

Xmodel-LM Technical Report