This research introduces a novel framework for selectively forgetting sensitive reasoning components in Large Reasoning Models without compromising their log...
Level: advanced
By Tuan Le
Category: research