There are a 1000 papers published on arXiv in ML every single day. It's hard to keep abreast of everything that's going on in this space.
ablation is not manual, and comes after evaluating which parts to turn off against given desired result.
it is manual, just automated via running inferences on a dataset of assumed refusals, and is a small rank-1 weight edit that doesn't require fine-tuning or training. Perf hit is much smaller than baking in these refusals in the first place.
small _automated_ rank-1 edit is not manual edit, sorry. automated is the contrary of manual by definition (which is "thinking by itself", definitely not something done by hand (main in french, thus manual)).