Getting Space Right: Agentic Rewards for Image Generation
cs.CV, cs.AI
Submitted: 2026-10-10
Updated: 2026-10-10
Sources
- Qwen3-VL Technical Report
- Qwen2.5-VL Technical Report
- Seed2.0 Model Card: Towards Intelligence Frontier for Real-World Complexity
- Scaling Spatial Intelligence with Multimodal Foundation Models
- DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
- DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression
- Emerging Properties in Unified Multimodal Pretraining
- Scaling Rectified Flow Transformers for High-Resolution Image Synthesis
- Large-Scale Universal Defect Generation: Foundation Models and Datasets
- Z-Reward: Beyond Scalar Rewards by Internalizing Reasoning into Score Distributions
- GenEval 2: Addressing Benchmark Drift in Text-to-Image Evaluation
- SpatialLadder: Progressive Training for Spatial Reasoning in Vision-Language Models
- HiSpatial: Taming Hierarchical 3D Spatial Understanding in Vision-Language Models
- Flow-GRPO: Training Flow Matching Models via Online RL
- Improving Video Generation with Human Feedback
- HPSv3++: Scaling Reward Models Across the Full Spectrum of Diffusion Model Capabilities
- Disentangling Pictorial Cue Understanding from Language Bias in VLMs via Depth Ordering Task
- XEmbodied: A Foundation Model with Enhanced Geometric and Physical Cues for Large-Scale Embodied Environments
- On the Design of Qwen3.8-Next Architecture: Evaluation, Efficiency, and Training Stability
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models