arXiv:2505.01618v3 Announce Type: replace Abstract: We study compute efficiency of LLM training when using different parameterizations, i.e., rules for adjusting model and optimizer hyperparameters (HPs) as model size changes. Some parameterizations fail to transfer optimal base HPs (such as learning…
arXiv:2505.16690v4 Announce Type: replace Abstract: Post-training of large language models is essential for adapting pre-trained language models (PLMs) to align with human preferences and downstream tasks. While PLMs typically exhibit well-calibrated confidence, post-trained language models (PoLMs) often suffer from over-confidence,…
arXiv:2510.17697v2 Announce Type: replace-cross Abstract: Steering cooperative multi-agent reinforcement learning (MARL) towards desired outcomes is challenging, particularly when the global guidance from a human on the whole multi-agent system is impractical in a large-scale MARL. On the other hand, designing…
arXiv:2502.03393v5 Announce Type: replace Abstract: Accurate epidemic forecasting is crucial for outbreak preparedness, but existing data-driven models are often brittle. Typically trained on a single pathogen, they struggle with data scarcity during new outbreaks and fail under distribution shifts caused…
arXiv:2408.05854v5 Announce Type: replace-cross Abstract: Goodness-of-fit testing is often criticized for its lack of practical relevance: since “all models are wrong”, the null hypothesis that the data conform to our model is ultimately always rejected as the sample size grows.…
arXiv:2505.24161v2 Announce Type: replace-cross Abstract: Spiking Neural Networks (SNNs) offer low-latency and energy-efficient decision making on neuromorphic hardware, making them attractive for Reinforcement Learning (RL) in resource-constrained edge devices. However, most RL algorithms for continuous control are designed for Artificial…
arXiv:2510.19933v1 Announce Type: new Abstract: The Muon optimizer has rapidly emerged as a powerful, geometry-aware alternative to AdamW, demonstrating strong performance in large-scale training of neural networks. However, a critical theory-practice disconnect exists: Muon’s efficiency relies on fast, approximate orthogonalization,…
arXiv:2510.19934v1 Announce Type: new Abstract: Differentially private (DP) decentralized Federated Learning (FL) allows local users to collaborate without sharing their data with a central server. However, accurately quantifying the privacy budget of private FL algorithms is challenging due to the…
arXiv:2510.19896v1 Announce Type: new Abstract: In this paper, we propose an approach to support the diagnosis of urinary tract diseases, with a focus on bladder cancer, using SHAP (SHapley Additive exPlanations)-based feature selection to enhance the transparency and effectiveness of…
arXiv:2510.19917v1 Announce Type: new Abstract: ”Noisy” datasets (regimes with low signal to noise ratios, small sample sizes, faulty data collection, etc) remain a key research frontier for classification methods with both theoretical and practical implications. We introduce FINDER, a rigorous…
