6 papers
Who Should We Listen to More? Welfare-Aware Preference Acquisition for Pluralistic Alignment
Ruomeng Ding, Tianwei Gao, Lianrui Geng +2
Pluralistic alignment needs to accommodate diverse preferences under limited human feedback. Using AI simulators to label preference comparisons reduces annotation costs, but varia…
Agentic Data Engineering for LLMs: System Design and a Controlled Empirical Study
Ruomeng Ding, Qianli Shen, Zhaoyang Han +3
LLM data engineering requires iterative experimentation to determine how data selection, rewards, and curricula affect model learning, making it a natural setting for AI-for-AI (AI…
Speculative Evaluation of Stochastic LLMs
Qianli Shen, Xiang Li, Ruomeng Ding +3
Evaluating a stochastic large language model is costly: benchmark scores estimate expected performance from randomized rollouts, yet uniform repetition ignores sharp differences in…
Rubrics as an Attack Surface: Stealthy Preference Drift in LLM Judges
Ruomeng Ding, Yifei Pang, He Sun +3
Evaluation and alignment pipelines for large language models increasingly rely on LLM-based judges, whose behavior is guided by natural-language rubrics and validated on benchmarks…
Whom to Query for What: Adaptive Group Elicitation via Multi-Turn LLM Interactions
Ruomeng Ding, Tianwei Gao, Thomas P. Zollo +3
Eliciting information to reduce uncertainty about latent group-level properties from surveys and other collective assessments requires allocating limited questioning effort under r…
Regularizing Hidden States Enables Learning Generalizable Reward Model for LLMs
Rui Yang, Ruomeng Ding, Yong Lin +2
Reward models trained on human preference data have been proven to effectively align Large Language Models (LLMs) with human intent within the framework of reinforcement learning f…