提出四象限难度分类法，揭示模型学习难点需依赖任务相关信号。
Four Quadrants of Difficulty: A Simple Categorisation and its Limits
- 按人类/模型、通用/特定任务划分难度信号，构建四象限框架
- 仅任务相关特征与模型困难度显著对齐，通用特征独立无效
- 呼吁开发轻量级任务特异性难度估计器，提升训练效率
课程学习（Curriculum Learning, CL）通过估计样本难度并合理调度训练顺序来提升模型性能。在自然语言处理领域，难度常通过与任务无关的语言学启发式或人类直觉估算，隐含假设这些信号与神经网络实际学习难度相关。本文提出一个四象限难度信号分类体系：人类判断 vs. 模型感知，以及任务无关 vs. 任务相关。在自然语言理解数据集上系统分析发现，任务无关特征表现独立，而仅任务相关特征能与模型学习难度对齐。这一结果挑战了现有课程学习的常见直觉，强调需要设计轻量级、任务相关的难度估计方法，以更真实反映模型的学习动态。
原文摘要 · Abstract (English)
Curriculum Learning (CL) aims to improve the outcome of model training by estimating the difficulty of samples and scheduling them accordingly. In NLP, difficulty is commonly approximated using task-agnostic linguistic heuristics or human intuition, implicitly assuming that these signals correlate with what neural models find difficult to learn. We propose a four-quadrant categorisation of difficulty signals -- human vs. model and task-agnostic vs. task-dependent -- and systematically analyse their interactions on a natural language understanding dataset. We find that task-agnostic features behave largely independently and that only task-dependent features align. These findings challenge common CL intuitions and highlight the need for lightweight, task-dependent difficulty estimators that better reflect model learning behaviour.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。