theoretical insights
Understanding derived from the theoretical study of algorithms, models, or concepts in AI, often leading to new methods or improvements in existing techniques. These insights help advance the field by establishing foundational principles.
- Advancing Wasserstein Convergence Analysis of Score-Based Models: Insights from Discretization and Second-Order Acceleration
- Analog In-memory Training on General Non-ideal Resistive Elements: The Impact of Response Functions
- Breaking AR’s Sampling Bottleneck: Provable Acceleration via Diffusion Language Models
- Collective Counterfactual Explanations: Balancing Individual Goals and Collective Dynamics
- Continual Multimodal Contrastive Learning
- Double Descent Meets Out-of-Distribution Detection: Theoretical Insights and Empirical Analysis on the Role of Model Complexity
- Evolving and Regularizing Meta-Environment Learner for Fine-Grained Few-Shot Class-Incremental Learning
- How Ensembles of Distilled Policies Improve Generalisation in Reinforcement Learning
- HybridNorm: Towards Stable and Efficient Transformer Training via Hybrid Normalization
- Improving LLM General Preference Alignment via Optimistic Online Mirror Descent
- In Search of Adam’s Secret Sauce
- In Search of Adam’s Secret Sauce
- Learning to Specialize: Joint Gating-Expert Training for Adaptive MoEs in Decentralized Settings
- Model Reconciliation via Cost-Optimal Explanations in Probabilistic Logic Programming
- Projecting Assumptions: The Duality Between Sparse Autoencoders and Concept Geometry
- Rethinking PCA Through Duality
- Tight analyses of first-order methods with error feedback
- TranSUN: A Preemptive Paradigm to Eradicate Retransformation Bias Intrinsically from Regression Models in Recommender Systems