AI Paper Atlas

Research map / Large language models / Alignment and preferences

ProPerSim: Developing Proactive and Personalized AI Assistants through User-Assistant Simulation

Jiho Kim, Junseong Choi, Woosog Chay, Daeun Kyung, Yeonsu Kwon, Yohan Jo, Edward Choi

ICLR 2026 · Large language models › Alignment and preferences · approach: LLM-as-judge and feedback data

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add DPO preference training on user-agent ratings (ablation)ProPerAssistant vs ARS Memory · ProPerSim, 32 personasavg score (0-4)Improved
put reward scores in the prompt (ARS Memory) instead of omitting them (ablation)AR Memory · ProPerSim, 32 personasavg score (0-4)Improved
add action-recommendation history without scores (AR Memory) to No Memory (ablation)No Memory · ProPerSim, 32 personasavg score (0-4)No clear effect
train with DPO over 14 daysProPerAssistant · ProPerSim, 32 personassuccessful intervention rate (%): 51.06 -> 71.51Improved
train with DPO over 14 daysProPerAssistant · ProPerSim, 32 personasavg score (0-4): 2.2 -> 3.3Improved
train with DPO over 14 daysProPerAssistant · ProPerSim, 32 personasrecommendations per hour: 24 -> 6Improved
train with DPO over 14 daysProPerAssistant · ProPerSim, persona 3 (complex preferences)avg score (0-4): 1.93 -> 2.53Mixed
train with DPO over 14 daysProPerAssistant · Personal Preference criterioncriterion score: 0.77 -> 0.83 on recommended actionsNo clear effect
target personas low in extraversionProPerAssistant · ProPerSim Big Five analysisscore improvementMixed
target personas low in opennessProPerAssistant · ProPerSim Big Five analysisscore improvementMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy