中文
相关论文

相关论文: Arctic-SnowCoder: Demystifying High-Quality Data i…

200 篇论文

Recent studies show that in supervised fine-tuning (SFT) of large language models (LLMs), data quality matters more than quantity. While most data cleaning methods concentrate on filtering entire samples, the quality of individual tokens…

计算与语言 · 计算机科学 2026-03-12 Jinlong Pang , Na Di , Zhaowei Zhu , Jiaheng Wei , Hao Cheng , Chen Qian , Yang Liu

Large Language Models (LLMs) have shown significant promise in automated theorem proving, yet progress is often constrained by the scarcity of diverse and high-quality formal language data. To address this issue, we introduce…

计算与语言 · 计算机科学 2025-12-02 Xinyuan Zhou , Yi Lei , Xiaoyu Zhou , Jingyi Sun , Yu Zhu , Zhongyi Ye , Weitai Zhang , Quan Liu , Si Wei , Cong Liu

Synthetic data is a standard component in training large language models, yet systematic comparisons across design dimensions, including rephrasing strategy, generator model, and source data, remain absent. We conduct extensive controlled…

Large language models (LLMs) are routinely pre-trained on billions of tokens, only to start the process over again once new data becomes available. A much more efficient solution is to continually pre-train these models, saving significant…

Much of software-engineering research relies on the naturalness of code, the fact that code, in small code snippets, is repetitive and can be predicted using statistical language models like n-gram. Although powerful, training such models…

软件工程 · 计算机科学 2022-08-15 Ahmed Khanfir , Matthieu Jimenez , Mike Papadakis , Yves Le Traon

Pre-trained code models lead the era of code intelligence, with multiple models designed with impressive performance. However, one important problem, data augmentation for code data that automatically helps developers prepare training data…

软件工程 · 计算机科学 2026-01-29 Zeming Dong , Qiang Hu , Xiaofei Xie , Maxime Cordy , Mike Papadakis , Yves Le Traon , Jianjun Zhao

Foundation models and self-supervised learning (SSL) have become central to modern AI, yet research in this area remains hindered by complex codebases, redundant re-implementations, and the heavy engineering burden of scaling experiments.…

软件工程 · 计算机科学 2025-11-26 Randall Balestriero , Hugues Van Assel , Sami BuGhanem , Lucas Maes

The current pretraining paradigm for large language models relies on massive compute and internet-scale raw text, creating a significant barrier to foundational research. In contrast, biological systems demonstrate highly sample-efficient…

计算与语言 · 计算机科学 2026-05-21 Guan Wang , Changling Liu , Chenyu Wang , Cai Zhou , Yuhao Sun , Yifei Wu , Shuai Zhen , Luca Scimeca , Yasin Abbasi Yadkori

Large Language Models (LLMs) excel in general tasks, but adapting them to specialized domains relies on high-quality supervised fine-tuning (SFT) data. Although existing methods can identify subsets of high-quality data and reduce training…

计算与语言 · 计算机科学 2025-10-17 Zhaoyang Shang , Sibo Wei , Jianbin Guo , Rui Zhou , Lifeng Dong , Yin Luo

Code pre-trained models (CodePTMs) have recently demonstrated significant success in code intelligence. To interpret these models, some probing methods have been applied. However, these methods fail to consider the inherent characteristics…

软件工程 · 计算机科学 2022-12-13 Nuo Chen , Qiushi Sun , Renyu Zhu , Xiang Li , Xuesong Lu , Ming Gao

With the great success of pre-trained models, the pretrain-then-finetune paradigm has been widely adopted on downstream tasks for source code understanding. However, compared to costly training a large-scale model from scratch, how to…

软件工程 · 计算机科学 2022-03-16 Deze Wang , Zhouyang Jia , Shanshan Li , Yue Yu , Yun Xiong , Wei Dong , Xiangke Liao

Code generation is crucial in software engineering for automating the coding process efficiently. While test-time computation methods show promise, they suffer from high latency due to multiple computation rounds. To overcome this, we…

软件工程 · 计算机科学 2025-05-28 Xiaoqing Zhang , Yuhan Liu , Flood Sung , Xiuying Chen , Shuo Shang , Rui Yan

Previous language model pre-training methods have uniformly applied a next-token prediction loss to all training tokens. Challenging this norm, we posit that "9l training". Our initial analysis examines token-level training dynamics of…

计算与语言 · 计算机科学 2025-01-09 Zhenghao Lin , Zhibin Gou , Yeyun Gong , Xiao Liu , Yelong Shen , Ruochen Xu , Chen Lin , Yujiu Yang , Jian Jiao , Nan Duan , Weizhu Chen

Language models achieve impressive performance on a variety of knowledge, language, and reasoning tasks due to the scale and diversity of pretraining data available. The standard training recipe is a two-stage paradigm: pretraining first on…

Competitive programming poses a significant challenge for Code LLMs. While recent models have shown promise, they heavily rely on finite real-world data, raising concerns about scalability and contamination. In this paper, we investigate a…

计算与语言 · 计算机科学 2026-02-03 Jie Wu , Haoling Li , Xin Zhang , Jiani Guo , Jane Luo , Steven Liu , Yangyu Huang , Ruihang Chu , Scarlett Li , Yujiu Yang

Subword tokenization introduces a computational layer in language models where many distinct token sequences decode to the same surface form and preserve meaning, yet induce different internal computations. Despite this non-uniqueness,…

计算与语言 · 计算机科学 2026-01-14 Adrian Cosma , Stefan Ruseti , Emilian Radoi , Mihai Dascalu

Can a small amount of verified goal information steer the expensive self-supervised pretraining of foundation models? Standard pretraining optimizes a fixed proxy objective (e.g., next-token prediction), which can misallocate compute away…

机器学习 · 计算机科学 2026-01-30 Shuqi Ke , Giulia Fanti

The massive growth of self-supervised learning (SSL) has been witnessed in language, vision, speech, and audio domains over the past few years. While discrete label prediction is widely adopted for other modalities, the state-of-the-art…

音频与语音处理 · 电气工程与系统科学 2022-12-20 Sanyuan Chen , Yu Wu , Chengyi Wang , Shujie Liu , Daniel Tompkins , Zhuo Chen , Furu Wei

Pre-trained Large Language Models (LLMs) have shown success in a diverse set of language inference and understanding tasks. The pre-training stage of LLMs looks at a large corpus of raw textual data. The BabyLM shared task compares LLM…

计算与语言 · 计算机科学 2024-01-11 Khushi Bhardwaj , Raj Sanjay Shah , Sashank Varma

Current language modeling approaches are built around tokens. Text corpora are split into tokens, and models are trained by performing computations on these tokens, such as predicting the next token given the preceding ones as context. This…

计算与语言 · 计算机科学 2026-05-26 Elio Musacchio , Lucia Siciliani , Pierpaolo Basile