Optimization & Theory

Knowledge Distillation for Learned Image Compression ICCV 2025 Knowledge-Grounded Detection of Cryptocurrency Scams with Retrieval-Augmented LMs ACL 2025 KurTail : Kurtosis-based LLM Quantization EMNLP 2025 KVCrush: Key Value Cache size-reduction using similarity in head-behaviour ACML 2025 KV-Latent: Dimensional-level KV Cache Reduction with Frequency-aware Rotary Positional Embedding ACL 2025 L4Q: Parameter Efficient Quantization-Aware Fine-Tuning on Large Language Models ACL 2025 Label Set Optimization via Activation Distribution Kurtosis for Zero-Shot Classification with Generative Models EMNLP 2025 LaCoOT: Layer Collapse through Optimal Transport ICCV 2025 LADM: Long-context Training Data Selection with Attention-based Dependency Measurement for LLMs ACL 2025 LAMB: A Training-Free Method to Enhance the Long-Context Understanding of SSMs via Attention-Guided Token Filtering ACL 2025 Language Model Probabilities are Not Calibrated in Numeric Contexts ACL 2025 Language Models Grow Less Humanlike beyond Phase Transition ACL 2025 Language Models Resist Alignment: Evidence From Data Compression ACL 2025 LANTERN: Accelerating Visual Autoregressive Models with Relaxed Speculative Decoding ICLR 2025 Large Language Models Are Overparameterized Text Encoders NAACL 2025 Large Language Models Badly Generalize across Option Length, Problem Types, and Irrelevant Noun Replacements EMNLP 2025 Large Learning Rates Simultaneously Achieve Robustness to Spurious Correlations and Compressibility ICCV 2025 Large Vocabulary Size Improves Large Language Models ACL 2025 Last-iterate Convergence of Shuffling Momentum Gradient Method under the Kurdyka-Lojasiewicz Inequality JMLR 2025 LAVa: Layer-wise KV Cache Eviction with Dynamic Budget Allocation EMNLP 2025 LayerAct: Advanced Activation Mechanism for Robust Inference of CNNs AAAI 2025 Layer- and Timestep-Adaptive Differentiable Token Compression Ratios for Efficient Diffusion Transformers CVPR 2025 Layered-Parameter Perturbation for Zeroth-Order Optimization of Optical Neural Networks AAAI 2025 Layerwise Recurrent Router for Mixture-of-Experts ICLR 2025 LazyDiT: Lazy Learning for the Acceleration of Diffusion Transformers AAAI 2025