cs.AI updates on arXiv.org 07月22日 12:44
Attend or Perish: Benchmarking Attention in Algorithmic Reasoning
index_new5.html
../../../zaker_core/zaker_tpl_static/wap/tpl_guoji1.html

 

本文提出AttentionSpan算法基准,通过五个无限输入域的任务,评估Transformer模型在未见类型输入上的泛化能力和学习机制鲁棒性,揭示注意力机制在泛化失败中的作用。

arXiv:2503.01909v2 Announce Type: replace-cross Abstract: Can transformers learn to perform algorithmic tasks reliably across previously unseen input/output domains? While pre-trained language models show solid accuracy on benchmarks incorporating algorithmic reasoning, assessing the reliability of these results necessitates an ability to distinguish genuine algorithmic understanding from memorization. In this paper, we propose AttentionSpan, an algorithmic benchmark comprising five tasks of infinite input domains where we can disentangle and trace the correct, robust algorithm necessary for the task. This allows us to assess (i) models' ability to extrapolate to unseen types of inputs, including new lengths, value ranges or input domains, but also (ii)to assess the robustness of their learned mechanisms. By analyzing attention maps and performing targeted interventions, we show that attention mechanism directly causes failures in extrapolation. We make the implementation of all our tasks and interpretability methods publicly available at https://github.com/michalspiegel/AttentionSpan .

Fish AI Reader

Fish AI Reader

AI辅助创作,多种专业模板,深度分析,高质量内容生成。从观点提取到深度思考,FishAI为您提供全方位的创作支持。新版本引入自定义参数,让您的创作更加个性化和精准。

FishAI

FishAI

鱼阅,AI 时代的下一个智能信息助手,助你摆脱信息焦虑

联系邮箱 441953276@qq.com

相关标签

Transformer 算法可靠性 AttentionSpan 泛化能力 鲁棒性
相关文章