HaM jointly trains multiple LLM policies by maximizing the hypervolume of their objective vectors, yielding a diverse Pareto-covering set of behaviors without pre-specified human preferences.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Multi-Objective Alignment of Large Language Models Through Hypervolume Maximization
HaM jointly trains multiple LLM policies by maximizing the hypervolume of their objective vectors, yielding a diverse Pareto-covering set of behaviors without pre-specified human preferences.