A history-aware visually grounded critic for computer use agents
A test time intervention framework for long-horizon GUI agents.
The work introduces a test time intervention framework for long-horizon GUI agents. HiViG equips the agent with a goal progress oriented summary to produce better actions, and then critiques those actions with visually grounded reasoning to refine them. Across a diverse set of GUI environments, HiViG improves task success rates for open weight and Gemini-3-Flash by 5.8% / 9% relative to the strongest previous critics evaluated.
Latest publications
MEMGUARD: Preventing memory contamination in long-term memory-augmented large language models
A type-aware memory framework that preserves functional memory boundaries during memory construction and retrieval.
EMNLPSPARC-RAG: Adaptive Sequential–Parallel Scaling with Context Management for Retrieval-Augmented Generation
A multi-agent framework that coordinates sequential and parallel inference-time scaling.
EMNLPReadability reconsidered: A cross-dataset analysis of reference-free metric
An investigation of factors shaping human perceptions of text readability and comprehensibility.
EMNLP