言栖资讯收录 140252 篇 AI、开发与建站文章（English），按发布时间排序，支持搜索与按语言筛选。
arXiv:2611.07211v1 Announce Type: new Abstract: Looped transformers scale inference-time reasoning by repeatedly reusing a shared core, increasing computation w…
arXiv:2611.07198v1 Announce Type: new Abstract: Each new GPU generation is showcased by its peak instruction rates, yet users hardly ever run instructions direc…
arXiv:2611.07149v1 Announce Type: new Abstract: Attention-FFN disaggregation (AFD) separates attention from the experts of a mixture-of-experts (MoE) model and …
arXiv:2611.07105v1 Announce Type: new Abstract: For sufficiently large transformer-based Large Language Models (LLMs), whose parameters cannot be retained in on…
arXiv:2611.07084v1 Announce Type: new Abstract: Matrix optimizers such as Muon update whole weight matrices, which conflicts with the element- or row-wise shard…
arXiv:2606.07334v5 Announce Type: replace Abstract: This report treats chord-symbol sequences as an interpretable, controllable time series for genre-local harm…
arXiv:2605.04998v4 Announce Type: replace Abstract: This paper treats chord progression generation as a standalone task and asks how much old-domain data must b…
arXiv:2603.04122v2 Announce Type: replace Abstract: Audio Super-Resolution is a set of techniques aimed at high-quality estimation of the given signal as if it …
arXiv:2611.08130v1 Announce Type: cross Abstract: Music source separation extracts vocals and instruments from a mixed recording, enabling remixing, content pro…
arXiv:2611.08092v1 Announce Type: cross Abstract: Real-time voice agents need a speech recognizer that never waits for future audio and whose memory per call do…
arXiv:2611.08251v1 Announce Type: new Abstract: Recent advances in speech generation have spurred the development of speech editing techniques. However, achievi…
arXiv:2611.08011v1 Announce Type: new Abstract: Speech foundation models built on flow matching handle generation, editing, and processing in one system, but ev…
arXiv:2611.07423v1 Announce Type: new Abstract: Music production typically involves multiple editing operations that each require specialized tools and expertis…
arXiv:2611.07420v1 Announce Type: new Abstract: Accent conversion (AC) aims to transform accent-related acoustic characteristics of the source speech to a targe…
arXiv:2611.07384v1 Announce Type: new Abstract: Dance motion combines slowly varying body configurations with rapid local movements. We investigate how explicit…
arXiv:2611.07210v1 Announce Type: new Abstract: Neural audio autoencoders, a deep learning method for sound synthesis, are increasingly popular in AI-enhanced N…
arXiv:2611.06982v1 Announce Type: new Abstract: Chunk-interleaved full-duplex speech language models listen and speak in fixed time chunks, and each chunk must …
arXiv:2608.01134v3 Announce Type: replace Abstract: Total quadratic ball discrepancy measures the deviation of codeword counts from uniformity over all Hamming …
arXiv:2605.18099v2 Announce Type: replace Abstract: Secure uplink transmission in low-earth orbit (LEO) satellite constellations faces significant challenges du…
arXiv:2605.14823v4 Announce Type: replace Abstract: In this paper, a class of linear authentication codes with secrecy, which are equipped with simple encoding …