Share your thoughts, 1 month free Claude Pro on usSee more

Home/Benchmarks

Controllability on Harmless-helpful Group 1 (unseen)

1Kendall's Tau

MOC

Updated 12d ago

Evaluation Results

Method	Links
MOC 2026.04		1
RiC 2026.04		0.6

SOTA Paper

MOC

One Model for All: Multi-Objective Controllable Language Models

Dataset

Harmless-helpful

Follow for update

@wizwand_team Discord

Related Benchmarks

Multi-objective Alignment on HHH (Harmlessness, Helpfulness, Humor)HH-RLHF on HH-RLHF Multi-Objective Optimization on Helpful Assistant Humor-helpful Multi-Objective Optimization on Helpful Assistant Harmless-helpful Reward Model Controllability on Humor helpful

© 2026 wizwand

Blog Contact Changelog Swarm

Privacy Policy Terms of Service FAQs Swarm Docs