KI-Forschungspapiere Täglich

Eliminating Oversaturation and Artifacts of High Guidance Scales in Diffusion Models

Contrastive Localized Language-Image Pre-Training

Open-RAG: Enhanced Retrieval-Augmented Reasoning with Open-Source Large Language Models

Loong: Generating Minute-level Long Videos with Autoregressive Language Models

VinePPO: Unlocking RL Potential For LLM Reasoning Through Refined Credit Assignment

CLIP-MoE: Towards Building Mixture of Experts for CLIP with Diversified Multiplet Upcycling

Contextual Document Embeddings

Layer Swapping for Zero-Shot Cross-Lingual Transfer in Large Language Models

MedVisionLlama: Leveraging Pre-Trained Large Language Model Layers to Enhance Medical Image Segmentation

Synthio: Augmenting Small-Scale Audio Classification Datasets with Synthetic Data

SciPrompt: Knowledge-augmented Prompting for Fine-grained Categorization of Scientific Topics

L-CiteEval: Do Long-Context Models Truly Leverage Context for Responding?

LLaVA-Critic: Learning to Evaluate Multimodal Models

MVGS: Multi-view-regulated Gaussian Splatting for Novel View Synthesis

Training Language Models on Synthetic Edit Sequences Improves Code Synthesis

Learning the Latent Rules of a Game from Data: A Chess Story

Distilling an End-to-End Voice Assistant Without Instruction Training Data

Robin3D: Improving 3D Large Language Model via Robust Instruction Tuning

Revisit Large-Scale Image-Caption Data in Pre-training Multimodal Foundation Models

Depth Pro: Sharp Monocular Metric Depth in Less Than a Second

Large Language Models as Markov Chains

Improving Autonomous AI Agents with Reflective Tree Search and Self-Learning

Intelligence at the Edge of Chaos

Interpreting and Editing Vision-Language Representations to Mitigate Hallucinations

Vinoground: Scrutinizing LMMs over Dense Temporal Reasoning with Short Videos

Video Instruction Tuning With Synthetic Data

SageAttention: Accurate 8-Bit Attention for Plug-and-play Inference Acceleration