Abstract

Large language models (LLM) have shown exceptional performance on a variety of natural language tasks. Yet, their capabilities for HTML understanding – i.e., parsing the raw HTML of a webpage, with applications to automation of web-based tasks, crawling, and browser-assisted retrieval – have not been fully explored. We contribute HTML understanding models (fine-tuned LLMs) and an in-depth analysis of their capabilities under three tasks: (i) Semantic Classification of HTML elements, (ii) Description Generation for HTML inputs, and (iii) Autonomous Web Navigation of HTML pages. While previous work has developed dedicated architectures and training procedures for HTML understanding, we show that LLMs pretrained on standard natural language corpora transfer remarkably well to HTML understanding tasks. For instance, fine-tuned LLMs are 12% more accurate at semantic classification compared to models trained exclusively on the task dataset. Moreover, when fine-tuned on data from the MiniWoB benchmark, LLMs successfully complete 50% more tasks using 192x less data compared to the previous best supervised model. To promote further research on LLMs for HTML understanding, we create and open-source a large-scale HTML dataset distilled and auto-labeled from CommonCrawl. We show evidence that T5-based models due to the bidirectional encoder-decoder architecture are the best choice and that for practitioners larger models are not necessarily better.

Web Navigation Task Animations

We show sample web navigation task animations from T5 models trained under different configurations. The full set of gifs is available at this anonymous github repo.

T5-3B

This model (named WebN-T5-3B in the paper) uses pretrained T5 model and includes action history as part of the input.

T5-3B-No-History

Without action history as input, this model sometimes stucks in circles (e.g. select then unselect a checkbox).

T5-3B-No-Pretrain

Without pretraining, the model has a low success rate.