Research map / Label-efficient and robust learning / Recognition and long-tail
Importance-aware Co-teaching for Offline Model-based Optimization
Ye Yuan, Can Chen, Zixuan Liu, Willie Neiswanger, Xue Liu
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 |
|---|---|---|---|
| add co-teaching (ablation) | ICT · TF Bind 8 | 100th percentile normalized ground-truth score: 0.905 ± 0.042 -> 0.958 ± 0.008 | Improved |
| add sample reweighting (ablation) | ICT · TF Bind 8 | 100th percentile normalized ground-truth score: 0.910 ± 0.024 -> 0.958 ± 0.008 | Improved |
| add co-teaching (ablation) | ICT · TF Bind 10 | 100th percentile normalized ground-truth score: 0.653 ± 0.018 -> 0.691 ± 0.023 | Improved |
| add sample reweighting (ablation) | ICT · TF Bind 10 | 100th percentile normalized ground-truth score: 0.654 ± 0.023 -> 0.691 ± 0.023 | Improved |
| add co-teaching (ablation) | ICT · NAS | 100th percentile normalized ground-truth score: 0.779 ± 0.071 -> 0.667 ± 0.091 | Hurt |
| add sample reweighting (ablation) | ICT · NAS | 100th percentile normalized ground-truth score: 0.666 ± 0.090 -> 0.667 ± 0.091 | No clear effect |
| add co-teaching (ablation) | ICT · SuperC | 100th percentile normalized ground-truth score: 0.500 ± 0.017 -> 0.503 ± 0.017 | No clear effect |
| add sample reweighting (ablation) | ICT · SuperC | 100th percentile normalized ground-truth score: 0.501 ± 0.017 -> 0.503 ± 0.017 | No clear effect |
| add co-teaching (ablation) | ICT · Ant Morphology | 100th percentile normalized ground-truth score: 0.927 ± 0.033 -> 0.961 ± 0.007 | Improved |
| add sample reweighting (ablation) | ICT · Ant Morphology | 100th percentile normalized ground-truth score: 0.914 ± 0.015 -> 0.961 ± 0.007 | Improved |
| add co-teaching (ablation) | ICT · D'Kitty Morphology | 100th percentile normalized ground-truth score: 0.962 ± 0.021 -> 0.968 ± 0.020 | No clear effect |
| add sample reweighting (ablation) | ICT · D'Kitty Morphology | 100th percentile normalized ground-truth score: 0.959 ± 0.013 -> 0.968 ± 0.020 | No clear effect |
Closest papers by meaning
- Offline Model-Based Optimization via Policy-Guided Gradient Search (AAAI 2024 · similarity 0.88)
- Learning Surrogates for Offline Black-Box Optimization via Gradient Matching (ICML 2024 · similarity 0.86)
- High-Dimensional Bayesian Optimization via Semi-Supervised Learning with Optimized Unlabeled Data Sampling (ICML 2024 · similarity 0.86)
- Robust Model-Based Optimization for Challenging Fitness Landscapes (ICLR 2024 · similarity 0.85)
- Generative Adversarial Model-Based Optimization via Source Critic Regularization (NeurIPS 2024 · similarity 0.85)
- Bridging Model-Based Optimization and Generative Modeling via Conservative Fine-Tuning of Diffusion Models (NeurIPS 2024 · similarity 0.85)
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.
