Research map / Label-efficient and robust learning / Recognition and long-tail
Pre-Trained Model Reusability Evaluation for Small-Data Transfer Learning
Yao-Xiang Ding, Xi-Zhu Wu, Kun Zhou, Zhi-Hua Zhou
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 SynLearn instead of LEEP | model reusability evaluation · MiniImageNet -> CIFAR-100, K=5 | Kendall tau: 0.282 -> 0.459 | Improved |
| use SynLearn instead of LEEP | model reusability evaluation · CIFAR-100 -> MiniImageNet, K=5 | Kendall tau: 0.235 -> 0.419 | Improved |
| use SynLearn instead of LEEP | model reusability evaluation · MiniImageNet -> CIFAR-100, K=20 | weighted Kendall tau: 0.703 -> 0.750 | Improved |
| use SynLearn instead of LogME | model reusability evaluation · CIFAR-100 -> MiniImageNet, K=5 | weighted Kendall tau: 0.312 -> 0.620 | Improved |
| use SynLearn instead of NCE | model reusability evaluation · MiniImageNet -> CIFAR-100, K=10 | Kendall tau: 0.300 -> 0.482 | Improved |
| increase target instances per class K from 5 to 20 | LEEP · MiniImageNet -> CIFAR-100 | Kendall tau: 0.282 -> 0.441 | Improved |
| increase target instances per class K from 5 to 20 | SynLearn · MiniImageNet -> CIFAR-100 | Kendall tau: 0.459 -> 0.502 | Improved |
| increase target instances per class K from 5 to 20 | SynLearn · CIFAR-100 -> MiniImageNet | weighted Kendall tau: 0.620 -> 0.639 | No clear effect |
| increase target instances per class K from 5 to 20 | LogME · MiniImageNet -> CIFAR-100 | Kendall tau: 0.139 -> 0.255 | Improved |
| use direct MSE on target data instead of SynLearn with K=1 | SynLearn · dSprites in-dataset MRE | rank correlation of MRE score | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- LEEP — LEEP: A New Measure to Evaluate Transferability of Learned Representations (ICML 2020) · better on 12 of 12 shared results
- LogME — LogME: Practical Assessment of Pre-trained Models for Transfer Learning (ICML 2021) · better on 8 of 8 shared results
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
- A High-Dimensional Statistical Method for Optimizing Transfer Quantities in Multi-Source Transfer Learning (NeurIPS 2025 · similarity 0.83)
- Understanding the Transferability of Representations via Task-Relatedness (NeurIPS 2024 · similarity 0.83)
- Knowledge Transfer from Vision Foundation Models for Efficient Training of Small Task-specific Models (ICML 2024 · similarity 0.83)
- Provable Sample-Efficient Transfer Learning Conditional Diffusion Models via Representation Learning (NeurIPS 2025 · similarity 0.82)
- Evaluating Transferability in Retrieval Tasks: An Approach Using MMD and Kernel Methods (CVPR 2024 · similarity 0.82)
- H-ensemble: An Information Theoretic Approach to Reliable Few-Shot Multi-Source-Free Transfer (AAAI 2024 · similarity 0.82)
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.
