Procedural Memory Under Change: Reuse and Interference in Controlled Web Tasks
arXiv cs.AIen
arXiv cs.AI
AI Global WirearXiv:2609.09774v1 Announce Type: new Abstract: Procedural memory lets language agents reuse successful routines, but reuse presumes that a stored routine remains applicable. We study what happens when that presumption is deliberately violated. The study combines a retrospective, human-assisted interface-adaptation case from BrowserGym TimeWarp with controlled frozen-memory comparisons on synthetic shopping decisions. During the documented WebShop V1-V6 development path, interface-specific code was adapted while the separately stored high-level procedure was not reported to change; this phase does not constitute an autonomous memory-agent evaluation. In the controlled phase, an early pilot p
This is a short summary published by AI Global Wire. The full article is owned and hosted by arXiv cs.AI — open it there to read it in full.
Read the full story at arXiv cs.AI- Verktyg
- Forskning
- Agenter
- Företag
Related AI news
- Meta share price surges after personal AI agent Muse releaseEconomic Times Tech · September 10, 2026
- Exclusive: Bynario raises €2.1m to tackle cybersecurity’s AI slop problemSifted · September 10, 2026
- Donnerstag: Apples neue iPhones auch aufklappbar, KI-Agenten weiter ungezügeltheise online – KI · September 10, 2026
- Generative AI a new tool in Mali's information war: studyEconomic Times Tech · September 10, 2026
- Sources: Alibaba is set to lead a $300M round in AI model testing startup UniPat AI at a $2.5B valuation; UniPat founder Li Kuan worked at Alibaba's Tongyi Lab (Bloomberg)Techmeme · September 10, 2026
- OpenDiscoveryTrace: Process Traces for Evaluating AI Scientist WorkflowsarXiv cs.AI · September 10, 2026