KI-Forschungspapiere Täglich

Scene-Centric Unsupervised Panoptic Segmentation

JavisDiT: Joint Audio-Video Diffusion Transformer with Hierarchical Spatio-Temporal Prior Synchronization

SkyReels-A2: Compose Anything in Video Diffusion Transformers

Whisper-LM: Improving ASR Models with Language Models for Low-Resource Languages

Advances and Challenges in Foundation Agents: From Brain-Inspired Intelligence to Evolutionary, Collaborative, and Safe Systems

OpenCodeReasoning: Advancing Data Distillation for Competitive Coding

Audio-visual Controlled Video Diffusion with Masked Selective State Spaces Modeling for Natural Talking Head Generation

Interpreting Emergent Planning in Model-Free Reinforcement Learning

NeuralGS: Bridging Neural Fields and 3D Gaussian Splatting for Compact 3D Representations

WikiVideo: Article Generation from Multiple Videos

Instruction-Guided Autoregressive Neural Network Parameter Generation

Rethinking RL Scaling for Vision Language Models: A Transparent, From-Scratch Framework and Comprehensive Evaluation Scheme

Scaling Analysis of Interleaved Speech-Text Language Models

Efficient Model Selection for Time Series Forecasting via LLMs

ZClip: Adaptive Spike Mitigation for LLM Pre-Training

Scaling Laws in Scientific Discovery with AI and Robot Scientists

Envisioning Beyond the Pixels: Benchmarking Reasoning-Informed Visual Editing

Sparse Autoencoders Learn Monosemantic Features in Vision-Language Models

GPT-ImgEval: A Comprehensive Benchmark for Diagnosing GPT4o in Image Generation

Inference-Time Scaling for Generalist Reward Modeling

GenPRM: Scaling Test-Time Compute of Process Reward Models via Generative Reasoning

ShortV: Efficient Multimodal Large Language Models by Freezing Visual Tokens in Ineffective Layers

FreSca: Unveiling the Scaling Space in Diffusion Models