Quantization, distillation, pruning and serving work aimed at the same accuracy for less memory, latency and money.
arXiv:2605. 11165v3 Announce Type: replace Abstract: Federated learning (FL) in heterogeneous environments remains challenging because client models often differ in both architecture and data distribution.
arXiv:2607. 10666v1 Announce Type: cross Abstract: Deploying AI-based visual inspection in manufacturing is hard because requirements change often, new defect types appear, and large labeled datasets are rarely available.
arXiv:2607. 09767v1 Announce Type: cross Abstract: The paper presents a voice anonymization model focusing on preserving content rather than producing realistic speech.
arXiv:2607. 09999v1 Announce Type: cross Abstract: We show that post-training quantization can silently alter how large language models reason even when task accuracy is preserved.
arXiv:2607. 11473v1 Announce Type: cross Abstract: In this paper, we propose TECO, a multi-dimensional pruning framework to collaboratively prune the three dimensions (depth, width, and resolution) of convolutional neural networks (CNNs) for better execution efficiency on embedded hardware.
arXiv:2607. 10555v1 Announce Type: cross Abstract: Generative Large Language Models (LLMs) have revolutionized information retrieval, yet their strictly parametric nature frequently leads to severe factual hallucinations when confronted with complex queries beyond their epistemic boundaries.
arXiv:2607. 11257v1 Announce Type: cross Abstract: Pathology Foundation Models (PFMs) offer powerful Whole Slide Image (WSI) representations but suffer from massive computational costs.
arXiv:2607. 09842v1 Announce Type: new Abstract: We investigate whether identity-specifying system prompts produce statistically distinguishable geometric fingerprints in the hidden-state trajectories of four open-weight transformer language models spanning four post-training regimes: no training (Gemma-4-E4B base), multimodal RLHF (Gemma-4-E4B-it), RL distillation (DeepSeek-R1-Distill-Qwen-7B), and SFT (Qwen2.
arXiv:2607. 10461v1 Announce Type: cross Abstract: Layout-based 3D scene synthesizers place each object using two human-annotated channels: a categorical class label and a canonical-pose convention.
arXiv:2607. 09785v1 Announce Type: cross Abstract: Traditionally, continual learning has assumed access to labeled data, yet many real-world applications -- such as lifelong robotics -- require models to adapt continuously from unlabeled streams.
arXiv:2510. 11546v3 Announce Type: replace-cross Abstract: High-dimensional regression often suffers from heavy-tailed noise and outliers, which can severely undermine the reliability of least-squares based methods.
arXiv:2607. 10113v1 Announce Type: new Abstract: Large language model agents increasingly store reusable procedures outside the model.
arXiv:2607. 11211v1 Announce Type: new Abstract: The popularity of large language models (LLMs) escalates an ongoing demand for effective inference.
arXiv:2607. 11075v1 Announce Type: new Abstract: The choice of Modulation and Coding (MCS) type for a particular channel condition is made through link adaptation (LA) algorithms that operate at the MAC layer.
arXiv:2604. 16514v5 Announce Type: replace-cross Abstract: Autoregressive vision-language models (VLMs) deliver strong multimodal capability, but their token-by-token decoding imposes a fundamental inference bottleneck.
arXiv:2607. 09988v1 Announce Type: cross Abstract: Recommendation systems, from traditional multi-stage to recent unified generative architectures, face challenges in incorporating diverse contextual signals, such as trending topics, breaking news, cultural events, and cross-surface user activities, into their ranking pipelines.
arXiv:2607. 10836v1 Announce Type: new Abstract: Multi-agent ensembling multiplies active parameters and inference cost without answering three basic questions: which agents to consult, how deeply a query should traverse a hierarchy of agents, and when inter-agent communication is worth its cost.
arXiv:2410. 19553v2 Announce Type: replace-cross Abstract: This paper explores the impact of occlusions in video action detection.
arXiv:2607. 09696v1 Announce Type: cross Abstract: Deep neural models for click-through rate prediction often exhibit a sharp decline in validation performance immediately after the first training epoch despite continued improvement in training loss.
arXiv:2607. 10661v1 Announce Type: cross Abstract: Speculative decoding has significantly accelerated Large Language Model (LLM) inference by alleviating memory-bound bottlenecks.