论文标题
自我解剖代理的神经进化
Neuroevolution of Self-Interpretable Agents
论文作者
论文摘要
注意力不足是一种心理现象，它导致人们在视线中错过事情。这是在感知中选择性关注的结果，使我们能够专注于世界的重要部分，而不会分散无关的细节。通过选择性注意的动机，我们研究了通过自我注意瓶颈镜头感知世界的人造药物的特性。通过仅限制对视觉输入的一小部分的访问，我们表明它们的策略在像素空间中直接可以解释。我们发现神经进化非常适合培训基于视觉的增强学习（RL）任务的自我发挥体系结构，从而使我们能够合并可能包括离散的，非不同的操作的模块，这些操作对我们的代理商有用。我们认为，自我注意力具有与间接编码相似的属性，从某种意义上说，从少量的键争分参数中产生了大隐含权重矩阵，因此使我们的代理能够解决比现有方法少1000倍的基于挑战性的基于挑战的基于挑战的视觉任务。由于我们的代理只关注任务关键的视觉提示，因此它们能够概括为在传统方法失败时修改任务元素无关的环境。我们的结果和源代码的视频可在https://attentionagent.github.io/上获得
Inattentional blindness is the psychological phenomenon that causes one to miss things in plain sight. It is a consequence of the selective attention in perception that lets us remain focused on important parts of our world without distraction from irrelevant details. Motivated by selective attention, we study the properties of artificial agents that perceive the world through the lens of a self-attention bottleneck. By constraining access to only a small fraction of the visual input, we show that their policies are directly interpretable in pixel space. We find neuroevolution ideal for training self-attention architectures for vision-based reinforcement learning (RL) tasks, allowing us to incorporate modules that can include discrete, non-differentiable operations which are useful for our agent. We argue that self-attention has similar properties as indirect encoding, in the sense that large implicit weight matrices are generated from a small number of key-query parameters, thus enabling our agent to solve challenging vision based tasks with at least 1000x fewer parameters than existing methods. Since our agent attends to only task critical visual hints, they are able to generalize to environments where task irrelevant elements are modified while conventional methods fail. Videos of our results and source code available at https://attentionagent.github.io/
