Researchers · Interviewed at ICLR 2026

Smitha Milli

Meta

ICLR 2026 in Rio de Janeiro. Beyond RLHF: Learning Diverse Human Values with Community Alignment

ICLR 2026 · June 1, 2026 · 4 min

ICLR 2026 in Rio de Janeiro. Beyond RLHF: Learning Diverse Human Values with Community Alignment

Current LLM alignment methods assume we can learn human preferences from preference datasets. But what if those datasets are fundamentally missing the diversity of human values?

At #ICLR 2026, Meta researcher Smitha Milli presents the Community Alignment Dataset, a new open-source resource designed to capture how preferences vary across people, cultures, and communities.

More from ICLR 2026

All event interviews →