PhD Thesis: Exploring the role of (self-)attention in cognitive and computer vision architecture
We investigate the role of attention and memory in complex reasoning tasks. We analyze Transformer-based self-attention as a model and extend it with memory. By studying a synthetic visual reasoning test, we refine the taxonomy of reasoning tasks. Incorporating self-attention with ResNet50, we enhance feature maps using feature-based and spatial attention, achieving efficient solving of challenging visual reasoning tasks. Our findings contribute to understanding the attentional needs of SVRT tasks. Additionally, we propose GAMR, a cognitive architecture combining attention and memory, inspired by active vision theory. GAMR outperforms other architectures in sample efficiency, robustness, and compositionality, and shows zero-shot generalization on new reasoning tasks.
Code (0)
등록된 구현이 없습니다.
Tasks
Visual ReasoningZero-shot GeneralizationSimilar Papers 제목 키워드 기반
Cognitive Mirrors: Exploring the Diverse Functional Roles of Attention Heads in LLM Reasoning
Large language models (LLMs) have achieved state-of-the-art performance in a variety of tasks, but remain largely opaque in terms of their internal mechanisms. Understanding these mechanisms is crucial to improve their r…
Logical ReasoningThe role of temporal cortex in the control of attention
Attention is an indispensable component of active vision. Contrary to the widely accepted notion that temporal cortex processing primarily focusses on passive object recognition, a series of very recent studies emphasize…
Object RecognitionSTSBlind Spots and Biases: Exploring the Role of Annotator Cognitive Biases in NLP
With the rapid proliferation of artificial intelligence, there is growing concern over its potential to exacerbate existing biases and societal disparities and introduce novel ones. This issue has prompted widespread att…
Decision MakingSelf-Attention Limits Working Memory Capacity of Transformer-Based Models
Recent work on Transformer-based large language models (LLMs) has revealed striking limits in their working memory capacity, similar to what has been found in human behavioral studies. Specifically, these models' perform…
DecoderPositionExpectation and Locality Effects in the Prediction of Disfluent Fillers and Repairs in English Speech
This study examines the role of three influential theories of language processing, \textit{viz.}, Surprisal Theory, Uniform Information Density (UID) hypothesis and Dependency Locality Theory (DLT), in predicting disflue…