KI-Forschungspapiere Täglich

WritingBench: A Comprehensive Benchmark for Generative Writing

REF-VLM: Triplet-Based Referring Paradigm for Unified Visual Decoding

PE3R: Perception-Efficient 3D Reconstruction

DreamRelation: Relation-Centric Video Customization

PhiloBERTA: A Transformer-Based Cross-Lingual Analysis of Greek and Latin Lexicons

SEAP: Training-free Sparse Expert Activation Pruning Unlock the Brainpower of Large Language Models

SurveyForge: On the Outline Heuristics, Memory-Driven Generation, and Multi-dimensional Evaluation for Automated Survey Writing

A Data-Centric Revisit of Pre-Trained Vision Models for Robot Learning

LLaVE: Large Language and Vision Embedding Models with Hardness-Weighted Contrastive Learning

Next Token Is Enough: Realistic Image Quality and Aesthetic Scoring with Multimodal Large Language Model

DiffCLIP: Differential Attention Meets CLIP

EasyControl: Adding Efficient and Flexible Control for Diffusion Transformer

WISE: A World Knowledge-Informed Semantic Evaluation for Text-to-Image Generation

Seg-Zero: Reasoning-Chain Guided Segmentation via Cognitive Reinforcement

Vision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Models

Feynman-Kac Correctors in Diffusion: Annealing, Guidance, and Product of Experts

Zero-AVSR: Zero-Shot Audio-Visual Speech Recognition with LLMs by Learning Language-Agnostic Speech Representations

FEA-Bench: A Benchmark for Evaluating Repository-Level Code Generation for Feature Implementation

Escaping Plato's Cave: Towards the Alignment of 3D and Text Latent Spaces

Should VLMs be Pre-trained with Image Data?

Novel Object 6D Pose Estimation with a Single Reference View

Unleashing the Potential of Large Language Models for Text-to-Image Generation through Autoregressive Representation Alignment

Feature-Level Insights into Artificial Text Detection with Sparse Autoencoders

Effective and Efficient Masked Image Generation Models

AlphaDrive: Unleashing the Power of VLMs in Autonomous Driving via Reinforcement Learning and Reasoning

Promote, Suppress, Iterate: How Language Models Answer One-to-Many Factual Queries

Symbolic Mixture-of-Experts: Adaptive Skill-based Routing for Heterogeneous Reasoning

Beyond RAG: Task-Aware KV Cache Compression for Comprehensive Knowledge Reasoning

DistiLLM-2: A Contrastive Approach Boosts the Distillation of LLMs

This Is Your Doge, If It Please You: Exploring Deception and Robustness in Mixture of LLMs

What's in a Latent? Leveraging Diffusion Latent Space for Domain Generalization

NeuGrasp: Generalizable Neural Surface Reconstruction with Background Priors for Material-Agnostic Object Grasp Detection

Agent models: Internalizing Chain-of-Action Generation into Reasoning models

HumanMM: Global Human Motion Recovery from Multi-shot Videos

MM-Eureka: Exploring Visual Aha Moment with Rule-based Large-scale Reinforcement Learning

Taking Notes Brings Focus? Towards Multi-Turn Multimodal Dialogue Learning

FedRand: Enhancing Privacy in Federated Learning with Randomized LoRA Subparameter Updates

Words or Vision: Do Vision-Language Models Have Blind Faith in Text?

VACE: All-in-One Video Creation and Editing

State-offset Tuning: State-based Parameter-Efficient Fine-Tuning for State Space Models

Efficient Distillation of Classifier-Free Guidance using Adapters

Adaptive Audio-Visual Speech Recognition via Matryoshka-Based Multimodal LLMs

TRCE: Towards Reliable Malicious Concept Erasure in Text-to-Image Diffusion Models

RePO: ReLU-based Preference Optimization

YOLOE: Real-Time Seeing Anything

Automated Movie Generation via Multi-Agent CoT Planning

MedAgentsBench: Benchmarking Thinking Models and Agent Frameworks for Complex Medical Reasoning

Detection Avoidance Techniques for Large Language Models

ProBench: Judging Multimodal Foundation Models on Open-ended Multi-domain Expert Tasks

BlackGoose Rimer: Harnessing RWKV-7 as a Simple yet Superior Replacement for Transformers in Large-Scale Time Series Modeling