Research map / Learning theory and algorithms / Prediction and decision losses
How Hard Can It Be? Hardness-Aware Multi-Objective Unlearning
Jiangwei Chen, Xinyuan Niu, Rachael Hwee Ling Sim, Zhengyuan Liu, Nancy F. Chen, Bryan Kian Hsiang Low
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 layer-specific constraints instead of a global constraint (ablation) | HAMU-Q / HAMU-U · image classification unlearning | forget quality and retain utility | Improved |
| increase Q (ablation) | HAMU-Q · unlearning trajectories | forget quality vs retain utility | Mixed |
| increase U (ablation) | HAMU-U · unlearning trajectories | retain utility vs forget quality | Mixed |
| increase Q (ablation) | HAMU-Q · epochs to reach L_f=1.0 | number of epochs | Improved |
| increase similarity mixing ratio rho to 0.75 | baselines (GA, KL, FT, SCRUB) · forget/retain similarity | trajectory | Hurt |
| use HAMU instead of baselines at rho=0.75 | HAMU variants · hard unlearning scenario | delta L_f / -delta L_r | Improved |
| increase rho to 1 | HAMU-Q / HAMU-U · unlearning trajectories | delta L_f / -delta L_r | Hurt |
| make forget data semantically similar to retain data | HAMU · WaterDrum-TOFU | estimated hardness: 4.0e-4 -> 6.1e-4 | Hurt |
| use HAMU-Q instead of GDiff | HAMU-Q · WaterDrum-TOFU LLM QA | forget quality and retain utility | Improved |
| use smaller learning rate (ablation) | HAMU-U · WaterDrum-TOFU dissimilar | retain objective constraint | Improved |
Closest papers by meaning
- When to Forget? Complexity Trade-offs in Machine Unlearning (ICML 2025 · similarity 0.93)
- What makes unlearning hard and what to do about it (NeurIPS 2024 · similarity 0.91)
- Efficient Utility-Preserving Machine Unlearning with Implicit Gradient Surgery (NeurIPS 2025 · similarity 0.89)
- LoTUS: Large-Scale Machine Unlearning with a Taste of Uncertainty (CVPR 2025 · similarity 0.89)
- Towards Source-Free Machine Unlearning (CVPR 2025 · similarity 0.88)
- Machine Unlearning under Overparameterization (NeurIPS 2025 · similarity 0.88)
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.
