Human-verified, frozen memory improves computer-using agents' task success and blocks memory-poisoning attacks without additional fine-tuning.
Osworld: Benchmarking multimodal agents for open-ended tasks in real computer environments, 2024
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
dataset 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1roles
dataset 1polarities
support 1representative citing papers
citing papers explorer
-
VerificAgent: Domain-Specific Memory Verification for Scalable Oversight of Aligned Computer-Use Agents
Human-verified, frozen memory improves computer-using agents' task success and blocks memory-poisoning attacks without additional fine-tuning.