Text Classification on Overruling
72AccuracyLLM w/o cascade
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| LLM w/o cascadeCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct2026.05 | 72 | 21.42 | 6.67 | 100 | |
| InitialCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct2026.05 | 71 | 13.47 | 3.95 | 54 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=12026.05 | 67.5 | 19.46 | 6.1 | 85.5 | |
| Single-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=SLM only2026.05 | 66.5 | 17.51 | 5.61 | 74 | |
| Random noiseCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct2026.05 | 65.5 | 16.02 | 4.83 | 62.5 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=0.82026.05 | 61.5 | 16.94 | 5.27 | 70 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=0.62026.05 | 57 | 14.66 | 4.46 | 59.5 | |
| Single-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=DM only2026.05 | 48.5 | 12.59 | 3.19 | 47 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=0.42026.05 | 45.5 | 12.57 | 3.21 | 47 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=0.22026.05 | 37 | 7.94 | 2.86 | 29.5 | |
| Joint-target attackCascade system=GPT2-medium + BERT-base + Phi-3-medium-4k-instruct, Setting=P=02026.05 | 21 | 4.78 | 2.04 | 17.5 |