Text Classification on Headline
78Performance ScoreInitial
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| InitialCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct2026.05 | 78 | 13.43 | 4.8 | 55 | |
| LLM w/o cascadeCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct2026.05 | 76 | 25.23 | 7.12 | 100 | |
| Random noiseCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct2026.05 | 69.5 | 16.83 | 5.41 | 69 | |
| Single-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=SLM only2026.05 | 69 | 20.21 | 5.64 | 72 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=12026.05 | 63 | 22.45 | 6.56 | 84.5 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=0.82026.05 | 56.5 | 20.09 | 5.59 | 71 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=0.62026.05 | 50 | 18.34 | 4.78 | 62 | |
| Single-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=DM only2026.05 | 46.5 | 15.17 | 4.17 | 45 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=0.42026.05 | 43.5 | 15.99 | 4.04 | 46 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=0.22026.05 | 41 | 14.62 | 3.3 | 37.5 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=02026.05 | 23 | 7.26 | 2.47 | 21 |