Pith. sign in

arXiv preprint arXiv:2406.07657 , year=

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

fields

cs.CL 1

years

2026 1

verdicts

UNVERDICTED 1

representative citing papers

Steerable Cultural Preference Optimization of Reward Models

cs.CL · 2026-06-17 · unverdicted · novelty 5.0

SCPO is a steerable training method for reward models that improves minority cultural preference accuracy by up to 7 points and is up to 280% more data-efficient than standard finetuning on PRISM and GlobalOpinionQA datasets.

citing papers explorer

Showing 1 of 1 citing paper.

  • Steerable Cultural Preference Optimization of Reward Models cs.CL · 2026-06-17 · unverdicted · none · ref 23

    SCPO is a steerable training method for reward models that improves minority cultural preference accuracy by up to 7 points and is up to 280% more data-efficient than standard finetuning on PRISM and GlobalOpinionQA datasets.