WebServ delivers an efficient full-stack web environment for RL web agents, cutting container launch costs dramatically and enabling RL-trained 4B models to reach 55.5% on WebArena-Lite, beating larger baselines.
Prompting is Not All You Need! Evaluating LLM Agent Simulation Methodologies with Real-World Online Customer Behavior Data, June 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
WEBSERV: A Full-Stack and RL-Ready Web Environment for Training Web Agents at Scale
WebServ delivers an efficient full-stack web environment for RL web agents, cutting container launch costs dramatically and enabling RL-trained 4B models to reach 55.5% on WebArena-Lite, beating larger baselines.