error accumulation
The gradual build-up of errors in a model's predictions or computations over time or iterations, often leading to degradation in performance. This is particularly relevant in sequential decision-making tasks or dynamic environments.
- AC-DiT: Adaptive Coordination Diffusion Transformer for Mobile Manipulation
- AccuQuant: Simulating Multiple Denoising Steps for Quantizing Diffusion Models
- Balanced Conic Rectified Flow
- Corrector Sampling in Language Models
- DNAEdit: Direct Noise Alignment for Text-Guided Rectified Flow Editing
- Diffusion-Based Hierarchical Graph Neural Networks for Simulating Nonlinear Solid Mechanics
- Enhancing Consistency of Flow-Based Image Editing through Kalman Control
- Frame Context Packing and Drift Prevention in Next-Frame-Prediction Video Diffusion Models
- Hierarchical Implicit Neural Emulators
- MaNGO — Adaptable Graph Network Simulators via Meta-Learning
- Motion Matters: Compact Gaussian Streaming for Free-Viewpoint Video Reconstruction
- Non-Markovian Discrete Diffusion with Causal Language Models
- On the Entropy Calibration of Language Models
- ReservoirTTA: Prolonged Test-time Adaptation for Evolving and Recurring Domains
- Semi-Supervised Regression with Heteroscedastic Pseudo-Labels
- ToF-IP: Time-of-Flight Enhanced Sparse Inertial Poser for Real-time Human Motion Capture
- UniRelight: Learning Joint Decomposition and Synthesis for Video Relighting