Hacker News 熱門(buzzing.cc 中文翻譯)·· 13 天前AI 評分44
Next-Concept Modeling:快速並行的「由粗到細」文字生成
Next-Concept 建模:快速、并行的“从粗到细”文本生成
AI 導讀
Next-Concept Modeling(NCM)提出由粗到細的兩階段框架,以 Transformer 逐層生成離散編碼,並行預測每層所有編碼。
正文
Abstract:Diffusion language models (DLMs) offer parallel sampling for faster generation than autoregressive models, an increasingly important advantage as inference becomes a dominant computational workload. However, DLMs degrade quickly with fewer sampling steps, process the full sequence at every step, and require many passes over the dataset to reach good quality. We introduce Next-Concept Modeling (NCM), a two-stage, coarse-to-fine framework: a multiscale residual VQ-VAE encodes text into a hierarchy of discrete codes, and a transformer generates one level at a time, predicting all codes within a level in parallel. At matched parameter scales, NCM achieves better generative perplexity than the strongest DLM baseline on LM1B while running $21\times$ faster. On OpenWebText, NCM is competitive with the strongest DLM while running $10\times$ faster; on XSum, it matches the strongest DLM's ROUGE-L while being $31\times$ faster. On the reasoning task of Sudoku, NCM outperforms autoregressive and diffusion baselines across all difficulty levels, using fewer function evaluations. On the reasoning task of Sudoku, NCM outperforms autoregressive and diffusion baselines across all difficulty levels, using fewer function evaluations.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.14162 [cs.CL] |
| (or arXiv:2610.14162v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.14162 arXiv-issued DOI via DataCite |
Submission history
From: Hermann Kumbong [view email]
[v1]
Fri, 9 Oct 2026 17:49:45 UTC (352 KB)
來源:Hacker News 熱門(buzzing.cc 中文翻譯) · arxiv.org