测试激活值能否提升上下文学习样本选择，发现效果不佳。
Activation-Based Active Learning for In-Context Learning: Challenges and Insights
- 用MLP激活值指导上下文示例选择，尝试精细优化。
- 最大斯皮尔曼相关系数仅0.33，激活值与示例质量无关。
- 可能因特征叠加导致信号失真，适合研究模型内部机制者阅读。
深度主动学习曾被用于大语言模型的上下文样本选择，但未结合对变压器激活机制的最新理解。本文检验了模型激活值能否提供细粒度信号以优化上下文示例选取。我们对MLP激活值驱动的深度主动学习方法进行了迄今最全面的分析，涵盖不同注意力掩码策略在多种分类与生成数据集上的表现，使用Llama-3.2-3B和Qwen2.5-3B基础模型。然而，结果为负面：无论从大规模激活或前四阶矩视角观察，MLP输出均不与示例质量或任务性能相关。所有任务与模型下，绝对斯皮尔曼相关系数最高仅为0.33，表明此类激活基采样不应用于上下文学习。我们推测这可能源于超叠加现象——模型所表示的特征数量超过其维度，提示稀疏自编码器（SAEs）等方法或为未来可行方向。
原文摘要 · Abstract (English)
Deep active learning has previously been explored for LLM in-context sample selection, but not with methods that utilise recent advances in understanding of transformer activations. In this paper, we test the hypothesis that model activations could provide a fine-grained signal to optimise the selection of in-context examples. We present the most comprehensive analysis to date of MLP activation-based deep active learning methods applied to in-context learning, including how different attention masking strategies impact active learning across diverse classification and generative datasets, using both Llama-3.2-3B and Qwen2.5-3B base models. However, we find a negative result: MLP outputs, viewed through the lenses of massive activations or the first four moments, do not correlate with example quality or task performance. Specifically, the absolute Spearman correlation coefficient is at most 0.33 for all tasks and models we tested, showing that such activation-based sampling should not be used for in-context learning. We hypothesise that this may be due to superposition, whereby models represent more features than they have dimensionality, suggesting that methods like Sparse Autoencoders (SAEs) may be a promising future direction.
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。