arXiv:2602.03396v3 Announce Type: replace
Abstract: Proprietary large language models (LLMs) embody substantial economic value and are generally exposed only as black-box APIs, yet adversaries can still exploit their outputs to extract knowledge via d…
arXiv:2605.04993v1 Announce Type: new
Abstract: Accurate forecasting of electric vehicle (EV) charging demand is critical for grid stability, infrastructure planning, and real-time charging optimization. In this work, we study the problem of early pre…
arXiv:2605.03314v2 Announce Type: replace
Abstract: In single-stream autoregressive interfaces, the same tokens both update the model state and constitute an irreversible public commitment. This coupling creates a silence tax: additional deliberation …
arXiv:2605.04208v1 Announce Type: new
Abstract: Large language models (LLMs) have demonstrated impressive multilingual capabilities for well-resourced languages, yet their performance on low-resource African languages remains poorly understood and lar…
arXiv:2602.02924v2 Announce Type: replace
Abstract: Diffusion policy sampling enables reinforcement learning (RL) to represent multimodal action distributions beyond suboptimal unimodal Gaussian policies. However, existing diffusion-based RL methods p…
arXiv:2605.04180v1 Announce Type: new
Abstract: Large Language Models exhibit strong reasoning and semantic understanding capabilities but often hallucinate in domains that require expert knowledge, among which fabrications, the generation of factuall…
arXiv:2602.05890v2 Announce Type: replace
Abstract: Training reinforcement learning (RL) systems in real-world environments remains challenging due to noisy supervision and poor out-of-domain (OOD) generalization, especially in LLM post-training. Rece…
arXiv:2605.05023v1 Announce Type: new
Abstract: Efficient CUDA implementations of attention mechanisms are critical to modern deep learning systems, yet supporting diverse and evolving attention variants remains challenging. Existing frameworks and co…
arXiv:2605.04221v1 Announce Type: new
Abstract: Clinical named entity recognition from dental progress notes is challenging because documentation is highly unstructured, domain-specific, and often privacy-sensitive. We developed a locally deployable f…
arXiv:2605.04209v1 Announce Type: cross
Abstract: We present Sparse Backdoor, a supply-chain attack that plants a \emph{provably undetectable} backdoor in pre-trained image classifiers, including convolutional networks and Vision Transformers. The att…