Post #2437636
2026-04-27 17:36 UTC
@mxp@mastodon.acm.org @tomstoneham@dair-community.social I guess my intuition is that unless the reversal is a strict “undo” nothing about the results is a) surprising and b) limited to LLMs nor even c) meaningfully classed as an error…. so it’s really all down to the actual operations themselves, so it’s all in the detail (and if “add Rembrandt lighting” really is an example they used, I see problems)
Replies (1)
-
@mxp@mastodon.acm.org 2026-04-27 18:30
@UlrikeHahn@fediscience.org @tomstoneham@dair-community.social Well, I wrote that it’s not surprising ;-) The results probably aren’t as meaningful as they may appear at first glance, but I still think that the paper raises a number of interesting questions, even though not all of them explicitly…