MANCE: Manifold-Aware Concept Erasure
anselm · x · 2026-07-11
MANCE (Manifold Aware Concept Erasure) introduces a concept erasure method designed to remove specific concepts from a model's representations while preserving other information as much as possible.
The core idea is to constrain the editing process to the natural data manifold, minimizing damage to the rest of the representation space. The authors claim this method achieved state-of-the-art results across 119 different settings.
More from Research
- Proto-tools, a universal infrastructure layer for generative biology, now open source — jmschreiber91 · 2026-10-26
- New paper finds AI agents negotiate badly under uncertainty and break deals on purpose — xuanalogue · 2026-10-26
- MCP fault injection study: agent task scores drop 60% under realistic tool failures, scale doesn't help — dair_ai · 2026-10-26
- Study: AI reviewing AI misses most reward hacking; best models miss 25%+ of hacks — ronbodkin · 2026-10-26
- FLUX.3 Klein step-distilled to 4-8 steps with CFG baked in — Fit-Structure-6489 · 2026-10-26
- NLI alone rejects 5/6 new agent memories; a 0.45 similarity gate fixes it all — Alternative_Pin9598 · 2026-10-26