Learning Theory

The Alignment Problem from a Deep Learning Perspective ICLR 2024 The Base-Rate Effect on LLM Benchmark Performance: Disambiguating Test-Taking Strategies from Benchmark Performance EMNLP 2024 The Benefits of Reusing Batches for Gradient Descent in Two-Layer Networks: Breaking the Curse of Information and Leap Exponents ICML 2024 The Best Arm Evades: Near-optimal Multi-pass Streaming Lower Bounds for Pure Exploration in Multi-armed Bandits COLT 2024 The Closeness of In-Context Learning and Weight Shifting for Softmax Regression NeurIPS 2024 The complexity of approximate (coarse) correlated equilibrium for incomplete information games COLT 2024 The Dimension of Self-Directed Learning ALT 2024 The Effective Horizon Explains Deep RL Performance in Stochastic Environments ICLR 2024 The Effect of Data Partitioning Strategy on Model Generalizability: A Case Study of Morphological Segmentation NAACL 2024 The Effect of Generalisation on the Inadequacy of the Mode EACL 2024 The Effect of Intrinsic Dataset Properties on Generalization: Unraveling Learning Differences Between Natural and Medical Images ICLR 2024 The Expressive Capacity of State Space Models: A Formal Language Perspective NeurIPS 2024 The Feature Speed Formula: a flexible approach to scale hyper-parameters of deep neural networks NeurIPS 2024 The Heuristic Core: Understanding Subnetwork Generalization in Pretrained Language Models ACL 2024 The High Line: Exact Risk and Learning Rate Curves of Stochastic Adaptive Learning Rate Algorithms NeurIPS 2024 The Human-AI Substitution game: active learning from a strategic labeler ICLR 2024 The Impact of Geometric Complexity on Neural Collapse in Transfer Learning NeurIPS 2024 The Implicit Bias of Gradient Descent on Separable Multiclass Data NeurIPS 2024 The Implicit Bias of Gradient Descent toward Collaboration between Layers: A Dynamic Analysis of Multilayer Perceptions NeurIPS 2024 The Implicit Bias of Heterogeneity towards Invariance: A Study of Multi-Environment Matrix Sensing NeurIPS 2024 The Impossibility of Parallelizing Boosting ALT 2024 The Joint Effect of Task Similarity and Overparameterization on Catastrophic Forgetting — An Analytical Model ICLR 2024 The Limits and Potentials of Local SGD for Distributed Heterogeneous Learning with Intermittent Communication COLT 2024 The Limits of Differential Privacy in Online Learning NeurIPS 2024 The mechanistic basis of data dependence and abrupt learning in an in-context classification task ICLR 2024