optimization framework
An optimization framework in AI provides a structured approach for fine-tuning and adjusting model parameters to minimize loss functions during training, often employing techniques like gradient descent or evolutionary algorithms.
- Co-Reinforcement Learning for Unified Multimodal Understanding and Generation
- Elastic Robust Unlearning of Specific Knowledge in Large Language Models
- Gaussian-Augmented Physics Simulation and System Identification with Complex Colliders
- HeavyWater and SimplexWater: Distortion-free LLM Watermarks for Low-Entropy Distributions
- How Far Are We from Optimal Reasoning Efficiency?
- Inverse Methods for Missing Data Imputation
- LORE: Lagrangian-Optimized Robust Embeddings for Visual Encoders
- Learning Multi-Source and Robust Representations for Continual Learning
- NopeRoomGS: Indoor 3D Gaussian Splatting Optimization without Camera Pose Input
- Revisiting Orbital Minimization Method for Neural Operator Decomposition
- Robustly Learning Monotone Single-Index Models
- SDPGO: Efficient Self-Distillation Training Meets Proximal Gradient Optimization
- Self-Training with Dynamic Weighting for Robust Gradual Domain Adaptation
- Simplicity Prevails: Rethinking Negative Preference Optimization for LLM Unlearning
- SiriuS: Self-improving Multi-agent Systems via Bootstrapped Reasoning
- Structure Matters: Dynamic Policy Gradient
- Unlabeled Data Improves Fine-Grained Image Zero-shot Classification with Multimodal LLMs
- VASA-3D: Lifelike Audio-Driven Gaussian Head Avatars from a Single Image
- Visual Sync: Multi‑Camera Synchronization via Cross‑View Object Motion