← back to paper

Review history

arxiv: 2507.21084 · 2 revisions

Reviving Your MNEME: Predicting The Side Effects of LLM Unlearning and Fine-Tuning via Sparse Model Diffing

1. 2026-08-15 CONDITIONAL MODERATE v1.4.0-alltime-deepseek-medium novelty 5.0
   125044 ms 13376 in 14610 out 2026-08-15T19:24:19.108283+00:00
2. 2026-08-06 REJECT LOW v1.4.0-alltime-deepseek-medium novelty 5.0
   139871 ms 13385 in 14101 out 2026-08-06T23:38:05.249441+00:00