Neural Network Optimization

Softmax is not Enough (for Sharp Size Generalisation) ICML 2025 SPAM: Spike-Aware Adam with Momentum Reset for Stable LLM Training ICLR 2025 Sparse Training from Random Initialization: Aligning Lottery Ticket Masks using Weight Symmetry ICML 2025 Spectral Scaling Laws in Language Models: emphHow Effectively Do Feed-Forward Networks Use Their Latent Space? EMNLP 2025 Speculative Streaming: Efficient and Scalable Speculative Decoding with Multi-Stream Attention EMNLP 2025 SSE-SAM: Balancing Head and Tail Classes Gradually Through Stage-Wise SAM AAAI 2025 Stability and Generalization of Zeroth-Order Decentralized Stochastic Gradient Descent with Changing Topology AAAI 2025 Stabilizing Sharpness-Aware Minimization Through A Simple Renormalization Strategy JMLR 2025 Stanford MLab at SemEval-2025 Task 11: Track B–Emotion Intensity Detection SemEval 2025 Steering Knowledge Selection Behaviours in LLMs via SAE-Based Representation Engineering NAACL 2025 Steering Protein Language Models ICML 2025 Step-KTO: Optimizing Mathematical Reasoning through Stepwise Binary Feedback EMNLP 2025 StitchLLM: Serving LLMs, One Block at a Time ACL 2025 Stop Walking in Circles! Bailing Out Early in Projected Gradient Descent CVPR 2025 Structured Outputs in Prompt Engineering: Enhancing LLM Adaptability on Counterintuitive Instructions AACL 2025 Structured Pruning for Diverse Best-of-N Reasoning Optimization ACL 2025 STUN: Structured-Then-Unstructured Pruning for Scalable MoE Pruning ACL 2025 SuBiTO: Synopsis-based Training Optimization for Continuous Real-Time Neural Learning over Big Streaming Data AAAI 2025 Subnet-Aware Dynamic Supernet Training for Neural Architecture Search CVPR 2025 SURGEON: Memory-Adaptive Fully Test-Time Adaptation via Dynamic Activation Sparsity CVPR 2025 SwiftPrune: Hessian-Free Weight Pruning for Large Language Models EMNLP 2025 Synonym-unaware Fast Adversarial Training against Textual Adversarial Attacks NAACL 2025 Syntax-Aware Retrieval Augmentation for Neural Symbolic Regression EMNLP 2025 SYSTRAN @ IWSLT 2025 Low-resource track ACL 2025 Taking a Big Step: Large Learning Rates in Denoising Score Matching Prevent Memorization COLT 2025