Quantization, distillation, pruning and serving work aimed at the same accuracy for less memory, latency and money.
arXiv:2602. 04349v3 Announce Type: replace-cross Abstract: 3D editing has emerged as a critical research area to provide users with flexible control over 3D assets.
arXiv:2606. 26112v1 Announce Type: cross Abstract: Low-resource languages face a critical challenge in AI development: creating specialized conversational systems without access to massive training corpora.
arXiv:2512. 02652v2 Announce Type: replace-cross Abstract: Existing methods for expressive music performance rendering, a conditional generation task that aims to generate a human-like performance from a symbolic score, rely on supervised learning over small labeled datasets, which limits scaling of both data volume and model size, despite the availability of vast unlabeled music, as in vision and language.
arXiv:2606. 26595v1 Announce Type: new Abstract: Enhancing the analysis of service feedback is essential for public sector organizations, particularly tax administrations, where trust and compliance depend on fair and effective service delivery.
arXiv:2606. 27147v1 Announce Type: cross Abstract: Unlike diffusion-based models that operate in continuous latent spaces, autoregressive unified multimodal models produce images by sequentially predicting discretized visual tokens.
arXiv:2606. 26587v1 Announce Type: cross Abstract: Low-bit floating-point formats and semi-structured sparsity are increasingly supported by modern accelerators, yet combining them for LLM activation compression remains challenging: activations contain input-dependent outliers that dominate block scales in FP4 quantization, and directly applying N:M sparsity masks discards moderate values, coupling sparsification loss with quantization error.
arXiv:2606. 26650v1 Announce Type: cross Abstract: In this paper, we present CAT-Q, Cost-efficient and Accurate Ternary Quantization, for compressing and accelerating LLMs.
arXiv:2606. 26769v1 Announce Type: new Abstract: The adoption of powerful diffusion models is hindered by their significant inference latency.
arXiv:2601. 01701v2 Announce Type: replace-cross Abstract: Anomaly detection is increasingly becoming crucial for maintaining the safety, reliability, and efficiency of industrial systems.
arXiv:2512. 10388v3 Announce Type: replace-cross Abstract: Conventional Sequential Recommender Systems (SRS) typically assign unique hash IDs (HID) to construct item embeddings, which mainly capture collaborative signals from historical user-item interactions.
arXiv:2606. 26787v1 Announce Type: cross Abstract: Traditional dynamic pricing models in large-scale e-commerce suffer from limited interpretability, poor utilization of unstructured information, and misalignment with long-term business objectives such as cumulative Gross Merchandise Value (GMV), Return on Investment (ROI) and milestone achievement.
arXiv:2606. 26424v1 Announce Type: new Abstract: Trajectory forecasting for autonomous driving has advanced rapidly, yet representative models often produce uninformative posteriors over forecast modes, causing problems for mode pruning.
arXiv:2606. 26538v1 Announce Type: cross Abstract: Deep Transformers are composed of uniformly stacked residual blocks, yet their deepest layers often add little value.
arXiv:2606. 26749v1 Announce Type: new Abstract: Neural Collapse predicts that balanced one-hot classification pushes model representations to be equally far from each other; a symmetric configuration that depends only on the output label and ignores any semantic similarity in the inputs.
Unlike diffusion-based models that operate in continuous latent spaces, autoregressive unified multimodal models produce images by sequentially predicting discretized visual tokens. These tokens are derived from a codebook that maps embeddings to quantized visual patterns.
Deploying large language models (LLMs) on Industrial Internet of Things (IIoT) edge devices demands extreme compression, yet existing structured pruning methods collapse at high compression ratios due to one-shot importance estimation, and their cross-architecture behavior remains unpredictable. This article presents a cascaded multi-granularity pruning framework that removes layers, attention heads, and feed-forward channels in coarse-to-fine order, with lightweight low-rank recovery between stages to re-estimate component importance.
The adoption of powerful diffusion models is hindered by their significant inference latency. Recent ``cache-then-forecast'' schemes alleviate this issue by accelerating DiTs using derivative-based polynomials, but they suffer from severe quality degradation at high acceleration ratios.
Neural Collapse predicts that balanced one-hot classification pushes model representations to be equally far from each other; a symmetric configuration that depends only on the output label and ignores any semantic similarity in the inputs. This creates a puzzle: next-token prediction language models are trained predominantly (as context length increases) with one-hot labels: the same context is very unlikely to appear twice in training with different labels.
Agents often repeatedly solve similar task instances from scratch, leading to unnecessary reasoning cost and long execution traces. Prior work has explored workflow reuse and executable skill induction, but it remains unclear which task scenarios admit procedural skills and how the shared procedural structure should be represented across successful traces.
Autoregressive large language model (LLM) serving is increasingly limited by key-value (KV) cache movement rather than dense matrix multiplication. Modern paged-attention systems reduce KV-cache fragmentation and mature kernels such as FlashInfer provide highly optimized native-paged decode attention.