Research map / Learning theory and algorithms / Prediction and decision losses
Deep Supervised Summarization: Algorithm and Application to Learning Instructions
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 all three losses (medoid + inter + intra) versus medoid only (ablation) | SupFL(N) · ProceL | F1 score: 61.6 -> 67.8 | Helped |
| add all three losses versus inter-cluster only (ablation) | SupFL(N) · ProceL | F1 score: 59.9 -> 67.8 | Helped |
| add all three losses versus intra-cluster only (ablation) | SupFL(N) · ProceL | F1 score: 59.3 -> 67.8 | Helped |
| use inter-cluster + intra-cluster loss without medoid (ablation) | SupFL(N) · ProceL | F1 score: 67.8 -> 58.2 | Hurt |
| learn representation (SupFL(N)) versus unsupervised UFL | UFL · ProceL | F1 score: 62.7 -> 67.8 | Helped |
| learn representation (SupFL(N)) versus UFL on tie a tie | UFL · ProceL tie a tie | F1 score: 51.6 -> 60.0 | Helped |
| SupFL(N) versus FCSN | FCSN · ProceL | F1 score: 63.8 -> 67.8 | Helped |
| SupFL(N) versus dppLSTM | dppLSTM · ProceL | F1 score: 56.6 -> 67.8 | Helped |
| linear transform SupFL(L) versus nonlinear SupFL(N) on Breakfast | SupFL · Breakfast | F1 score: 65.7 -> 66.9 | Mixed |
| learned representation versus UFL on Breakfast | UFL · Breakfast | F1 score: 64.5 -> 66.9 | Helped |
| vary lambda from 0.001 to 0.1 | SupFL · ProceL | F1 score: at most 1.2% | No clear effect |
| vary rho_inter from 0.01 to 10 | SupFL · ProceL | F1 score: at most 0.6% | No clear effect |
Closest papers by meaning
- Video Summarization by Learning From Unpaired Data (CVPR 2019 · similarity 0.87)
- Online Summarization via Submodular and Convex Optimization (CVPR 2017 · similarity 0.87)
- Video Summarization Using Denoising Diffusion Probabilistic Model (AAAI 2025 · similarity 0.83)
- FedSum: Data-Efficient Federated Learning Under Data Scarcity Scenario for Text Summarization (AAAI 2025 · similarity 0.82)
- Enhancing Semi-Supervised Learning via Representative and Diverse Sample Selection (NeurIPS 2024 · similarity 0.82)
- Multinoulli Extension: A Lossless Yet Effective Probabilistic Framework for Subset Selection over Partition Constraints (ICML 2025 · 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.