ML Anthology

Ng, Him Wai

1 publications

ICLR 2026 Once-More: Continuous Self-Correction for Large Language Models via Perplexity-Guided Intervention Jiaxun Gao, Him Wai Ng, Z. Jane Wang