Get Started
Research questionHow can AI agents adapt to each user’s evolving quality criteria on open-ended tasks?On open-ended writing and visual-creation tasks, users often judge outputs by standards they cannot fully articulate in advance. These standards may emerge and shift during interaction, leaving broadly capable agents unable to consistently meet an individual’s professional expectations.
AI
AI Agents
Machine Learning
Latest papersRecent research connected to this question, newest first.Efficient Test-Time Adaptation through Human-AI InteractionThe evidence concerns adaptation for 30 individuals across writing and visual-creation tasks, covering 600 tasks. Interaction signals are incorporated into agent context and weights alongside an evolving rubric. The reported results show 4.5–20.9% gains in solo task success within tens of tasks, rubrics that identify 16.0–22.3% more failures than language models or humans alone, and up to 8.8% cross-user improvement.research paper · Sep 3, 2026
Related questions
How can personalized dialogue agents keep evolving user personas traceable to supporting events during retrieval?How can computer-use agents retain, refine, and reliably reuse procedural skills across repeated GUI tasks?How should AI assistants choose proposals when users cannot reliably evaluate them?How can reasoning models keep improving on open-ended agentic tasks as human supervision and reliable rewards recede?
Home
Topics
Search
Library