Multi-step ahead prediction in language models is challenging due to the discrepancy between training and test time processes. At test time, a sequence predictor is required to make predictions given past predictions as the input, instead of the past targets that are provided during training. This difference, known as exposure bias, can lead to the compounding of errors along a generated sequence at test time. To improve generalization in neural language models and address compounding errors, we propose \textit{Nearest-Neighbor Replacement Sampling} -- a curriculum learning-based method that gradually changes an initially deterministic teacher policy to a stochastic policy. A token at a given time-step is replaced with a sampled nearest neighbor of the past target with a truncated probability proportional to the cosine similarity between the original word and its top $k$ most similar words. This allows the learner to explore alternatives when the current policy provided by the teacher is sub-optimal or difficult to learn from. The proposed method is straightforward, online and requires little additional memory requirements. We report our findings on two language modelling benchmarks and find that the proposed method further improves performance when used in conjunction with scheduled sampling.
PCC: Paraphrasing with Bottom-k Sampling and Cyclic Learning for Curriculum Data Augmentation
The Chinese University of Hong Kong（香港中文大学）
课程数据增强（CDA）通过呈现难度从易到难逐渐增加的合成数据来改进神经模型。然而，传统的批评性话语分析只是简单地将词的扰动率作为难度度量，并且只经过一次测试。本文提出了\textbf{PCC}：\textbf{P} araphrasing with Bottom-k Sampling and \textbf{C} cyclic Learning for \textbf{C} urgmentation数据增强，这是一种新的基于释义的CDA框架，它利用文本释义相似度作为课程难度的度量。我们提出了一个具有语义感知的释义生成模块，该模块由三个部分组成：一个具有bottom-k采样的释义候选生成器，一个过滤机制和一个难度度量。我们还提出了一个循环的学习策略，多次通过的cumulus。提出了bottom-k采样方法来生成超硬实例。对Few-Shot文本分类和对话生成的实验结果表明，PCC优于竞争基准。人类评估和大量的案例研究表明，bottom-k采样有效地生成超硬实例，PCC显着提高了基线对话代理。
Does the Definition of Difficulty Matter? Scoring Functions and their Role for Curriculum Learning
Technical University of Munich（慕尼黑工业大学）；MCML – Munich Center for Machine Learning（慕尼黑机器学习中心）；Imperial College（帝国理工学院）；MDSI – Munich Data Science Institute（慕尼黑数据科学研究所）
Zero-Shot Dependency Parsing with Worst-Case Aware Automated Curriculum Learning
University of Copenhagen（哥本哈根大学）；Uppsala University（乌普萨拉大学）；KU Leuven（鲁汶大学）；National University of Defense Technology（国防科技大学）
大型多语言预训练语言模型，如mBERT和XLM-RoBERTa，被发现对于句法分析模型的跨语言迁移非常有效（Wu and Dredze 2019），但仅限于相关语言之间。然而，当解析真正的低资源语言时，源语言和训练语言很少相关。为了弥补这一差距，我们采用了一种多任务学习的方法，该方法依赖于自动化课程学习，以动态优化离群语言的解析性能。我们表明，这种方法是显着优于均匀和大小比例抽样的zero-shot设置。
Actor-Curator: Co-adaptive Curriculum Learning via Policy-Improvement Bandits for RL Post-Training
University of Illinois Chicago（伊利诺伊大学香槟分校）；Caltech（加州理工学院）；RPI（罗切斯特理工学院）；MBZUAI（澳门大学人工智能研究院）；University of Chicago（芝加哥大学）；NEC Laboratories America（NEC美国实验室）
SA-GCS: Semantic-Aware Gaussian Curriculum Scheduling for UAV Vision-Language Navigation
Sun Yat-Sen University（中山大学）；DP Technology（深势科技）；Beijing University Of Posts and Telecommunications（北京邮电大学）；Beihang University（北京航空航天大学）；Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）；XPENG（小鹏）；Chulalongkorn University（朱拉隆功大学）
