← Optimization & Theory
Deep Learning › Optimization & Theory ›

Theory

1072 directly classified papers

Papers per year

Papers

Towards Understanding Jailbreak Attacks in LLMs: A Representation Space Analysis EMNLP 2024
Unveiling Linguistic Regions in Large Language Models ACL 2024
On the Empirical Complexity of Reasoning and Planning in LLMs EMNLP 2024
Get rich quick: exact solutions reveal how unbalanced initializations promote rapid feature learning NIPS 2024
The Fine-Grained Complexity of Gradient Computation for Training Large Language Models NIPS 2024
Double-Descent Curves in Neural Networks: A New Perspective Using Gaussian Processes AAAI 2024
AboutMe: Using Self-Descriptions in Webpages to Document the Effects of English Pretraining Data Filters ACL 2024
Unraveling the Gradient Descent Dynamics of Transformers NIPS 2024
Epistemic Uncertainty Quantification For Pre-Trained Neural Networks CVPR 2024
Exponential Hardness of Optimization from the Locality in Quantum Neural Networks AAAI 2024
Scaling Laws Across Model Architectures: A Comparative Analysis of Dense and MoE Models in Large Language Models EMNLP 2024
Polyhedral Complex Derivation from Piecewise Trilinear Networks NIPS 2024
Nonlinear dynamics of localization in neural receptive fields NIPS 2024
A Phase Transition between Positional and Semantic Learning in a Solvable Model of Dot-Product Attention NIPS 2024
Understanding Surprising Generalization Phenomena in Deep Learning AAAI 2024
How does Gradient Descent Learn Features --- A Local Analysis for Regularized Two-Layer Neural Networks NIPS 2024
Towards Constituting Mathematical Structures for Learning to Optimize ICML 2023
Unveiling The Mask of Position-Information Pattern Through the Mist of Image Features ICML 2023
How Powerful are Shallow Neural Networks with Bandlimited Random Weights? ICML 2023
Minimum Width of Leaky-ReLU Neural Networks for Uniform Universal Approximation ICML 2023
On the Correctness of Automatic Differentiation for Neural Networks with Machine-Representable Parameters ICML 2023
FP-Diffusion: Improving Score-based Diffusion Models by Enforcing the Underlying Score Fokker-Planck Equation ICML 2023
Gradient Descent Monotonically Decreases the Sharpness of Gradient Flow Solutions in Scalar Networks and Beyond ICML 2023
Benign Overfitting in Two-layer ReLU Convolutional Neural Networks ICML 2023
Emergent Asymmetry of Precision and Recall for Measuring Fidelity and Diversity of Generative Models in High Dimensions ICML 2023