Research map / Optimization / Optimizers
OFMU: OPTIMIZATION-DRIVEN FRAMEWORK FOR MACHINE UNLEARNING
Sadia Asif, Mohammad Mohammadi Amiri
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use bi-level OFMU instead of gradient ascent | LLaMA-2-7B-chat · TOFU forget05 | model utility: 0.00 -> 0.65 | Improved |
| use OFMU instead of RMU | LLaMA-2-7B-chat · TOFU forget05 | model utility: 0.02 -> 0.65 | Improved |
| use OFMU instead of NPO | LLaMA-2-7B-chat · TOFU forget10 | model utility: 0.54 -> 0.61 | Improved |
| use OFMU instead of NPO | LLaMA-2-7B-chat · TOFU forget01 | forget quality: 0.40 -> 0.44 | No clear effect |
| use OFMU instead of NPO | LLaMA-3.2-1B-Instruct · TOFU forget10 | forget quality: 0.02 -> 0.41 | Improved |
| use OFMU instead of influence unlearning | CIFAR-10 model · CIFAR-10 class-wise | unlearning accuracy: 89.31 -> 81.51 | Hurt |
| use OFMU instead of influence unlearning | CIFAR-10 model · CIFAR-10 class-wise | MIA efficacy: 55.22 -> 59.76 | Improved |
| use OFMU instead of gradient ascent | CIFAR-10 model · CIFAR-10 random forgetting | unlearning accuracy: 0.78 -> 7.71 | Improved |
| use OFMU instead of RMU | LLaMA (WMDP) · WMDP Biosecurity | unlearning efficacy: 71.3 -> 72.8 | Improved |
| use OFMU instead of RMU | LLaMA (WMDP) · WMDP MMLU | utility accuracy: 74.2 -> 74.6 | Improved |
| use OFMU instead of NPO | LLaMA (WMDP) · WMDP Cybersecurity | unlearning efficacy: 64.9 -> 70.4 | Improved |
| use OFMU on CIFAR-10 random forgetting vs retraining | CIFAR-10 model · CIFAR-10 random forgetting | retain accuracy: 100.00 -> 92.25 | Hurt |
Closest papers by meaning
- Constrained Entropic Unlearning: A Primal-Dual Framework for Large Language Models (NeurIPS 2025 · similarity 0.92)
- Learning What to Forget: Improving LLM Unlearning via Learned Token-Level Importance (NeurIPS 2026 · similarity 0.92)
- Reversing the Forget-Retain Objectives: An Efficient LLM Unlearning Framework from Logit Difference (NeurIPS 2024 · similarity 0.91)
- SHRED: Retain-Set-Free Unlearning via Self-Distillation with Logit Demotion (NeurIPS 2026 · similarity 0.91)
- LLM Unlearning via Loss Adjustment with Only Forget Data (ICLR 2025 · similarity 0.91)
- When to Forget? Complexity Trade-offs in Machine Unlearning (ICML 2025 · similarity 0.90)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
