WebAgent combines task planning, HTML summarization via HTML-T5, and Python code synthesis via Flan-U-PaLM to improve real-world web task success by over 50% and achieve state-of-the-art results on MiniWoB and Mind2Web benchmarks.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2023 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A Real-World WebAgent with Planning, Long Context Understanding, and Program Synthesis
WebAgent combines task planning, HTML summarization via HTML-T5, and Python code synthesis via Flan-U-PaLM to improve real-world web task success by over 50% and achieve state-of-the-art results on MiniWoB and Mind2Web benchmarks.