Research map / Architectures and efficiency / RNNs and normalization
MAP: Low-compute Model Merging with Amortized Pareto Fronts via Quadratic Approximation
Lu Li, Tianyu Zhang, Zhiqi Bu, Suyuchen Wang, Huan He, Jie Fu, Yonghui Wu, Jiang Bian, Yong Chen, Yoshua Bengio
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 MAP instead of the strongest listed merging baseline | ViT-B/32 model merging · Zero-shot classification, 3–8 tasks | preference weighted sum of accuracies: 67.94 $\pm$ 1.79 -> \textbf{69.05$\pm$1.84} (3 tasks); 71.79$\pm$2.93 -> \textbf{72.84$\pm$4.4} (4 tasks); 76.84$\pm$4.7 | Improved |
| Use MAP instead of Task Arithmetic | ViT-B/32 model merging · Zero-shot classification, 2 tasks | preference weighted sum of accuracies: 70.73$\pm$1.84 -> 70.7$\pm$1.76 | No clear effect |
| Use MAP with 30 or 50 coefficient vectors instead of direct search | ViT-B/32 model merging · Zero-shot classification, 2 tasks | win rate over direct search: 49.81\% $(\pm 0.30)$ | Hurt |
| Use MAP with 30 or 50 coefficient vectors instead of direct search | ViT-B/32 model merging · Zero-shot classification, 3 tasks | win rate over direct search: 46.90\% $(\pm 0.71)$ | Hurt |
| Use MAP instead of sparse direct search | ViT-B/32 model merging · Zero-shot classification, 4–8 tasks | win rate over direct search: 50.67\% $(\pm 2.44)$ (4 tasks); 53.00\% $(\pm 1.88)$ (5 tasks); 60.71\% $(\pm 1.34)$ (6 tasks); 63.42\% $(\pm 1.91)$ (7 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- TIES-merging — TIES-Merging: Resolving Interference When Merging Models (NeurIPS 2023) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Pareto Merging: Multi-Objective Optimization for Preference-Aware Model Merging (ICML 2025 · similarity 0.88)
- CABS: Conflict-Aware and Balanced Sparsification for Enhancing Model Merging (ICML 2025 · similarity 0.88)
- Whoever Started the interference Should End It: Guiding Data-Free Model Merging via Task Vectors (ICML 2025 · similarity 0.88)
- Dynamic Model Merging Made Slim (NeurIPS 2026 · similarity 0.87)
- Twin-Merging: Dynamic Integration of Modular Expertise in Model Merging (NeurIPS 2024 · similarity 0.87)
- Modeling Multi-Task Model Merging as Adaptive Projective Gradient Descent (ICML 2025 · similarity 0.87)
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.
