Reflection-Based Memory For Web navigation Agents

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Azam, Ruhana, Vempaty, Aditya, Jagmohan, Ashish
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909634670886912
author Azam, Ruhana
Vempaty, Aditya
Jagmohan, Ashish
author_facet Azam, Ruhana
Vempaty, Aditya
Jagmohan, Ashish
contents Web navigation agents have made significant progress, yet current systems operate with no memory of past experiences -- leading to repeated mistakes and an inability to learn from previous interactions. We introduce Reflection-Augment Planning (ReAP), a web navigation system to leverage both successful and failed past experiences using self-reflections. Our method improves baseline results by 11 points overall and 29 points on previously failed tasks. These findings demonstrate that reflections can transfer to different web navigation tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2506_02158
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Reflection-Based Memory For Web navigation Agents
Azam, Ruhana
Vempaty, Aditya
Jagmohan, Ashish
Artificial Intelligence
Web navigation agents have made significant progress, yet current systems operate with no memory of past experiences -- leading to repeated mistakes and an inability to learn from previous interactions. We introduce Reflection-Augment Planning (ReAP), a web navigation system to leverage both successful and failed past experiences using self-reflections. Our method improves baseline results by 11 points overall and 29 points on previously failed tasks. These findings demonstrate that reflections can transfer to different web navigation tasks.
title Reflection-Based Memory For Web navigation Agents
topic Artificial Intelligence
url https://arxiv.org/abs/2506.02158