📝 "Conditional Language Policy: A General Framework for Steerable Multi-Objective Finetuning", to appear at EMNLP 2024.
X 🔗: x.com/kaiwenw_ai/status/1855304823760970056
Full Paper: arxiv.org/abs/2407.15762
Abstract: Reward-based finetuning is crucial for aligning language policies with intended behaviors (e.g., creativity and safety). A key challenge is to develop steerable language models that trade-off multiple (conflicting) objectives in a flexible and efficient manner. This paper presents Conditional Language Policy (CLP), a general framework for finetuning language models on multiple objectives. Building on techniques from multi-task training and parameter-efficient finetuning, CLP learn steerable models that effectively trade-off conflicting objectives at inference time. Notably, this does not require training or maintaining multiple models to achieve different trade-offs between the objectives. Through extensive experiments and ablations on two summarization datasets, we show that CLP learns steerable language models that outperform and Pareto-dominate the existing approaches for multi-objective finetuning.
Keywords: Reinforcement Learning, Multi-Objective Finetuning, Multi-task Learning, Parameter Efficient Training