Emergence of Episodic Memory in Transformers: Characterizing Changes in Temporal Structure of Attention Scores During Training
원문 초록
We investigate in-context temporal biases in attention heads and transformer outputs. Using cognitive science methodologies, we analyze attention scores and outputs of the GPT-2 models of varying sizes. Across attention heads, we observe effects characteristic of human episodic memory, including temporal contiguity, primacy and recency. Transformer outputs demonstrate a tendency toward in-context serial recall. Importantly, this effect is eliminated after the ablation of the induction heads, which are the driving force behind the contiguity effect. Our findings offer insights into how transformers organize information temporally during in-context learning, shedding light on their similarities and differences with human memory and learning.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Assay2Mol: large language model-based drug design using BioAssay context
Universal pre-training by iterated random computation
Evolving Prompts In-Context: An Open-ended, Self-replicating Perspective
Conquering the Retina: Bringing Visual in-Context Learning to OCT
Evaluating Cell Type Inference in Vision Language Models Under Varying Visual Context
PRO-V: An Efficient Program Generation Multi-Agent System for Automatic RTL Verification
토론
이 논문과 연결된 포럼 글이 아직 없습니다.