Multiple-choice Question Answering on MMLU T1
52.3AccuracyRandom Rehearsal
Evaluation results
| Method | Links | |
|---|---|---|
| Random RehearsalBackbone=LLaMA-3.1-Instruct-8B, Number of random seeds=32026.04 | 52.3 | |
| EWCBackbone=LLaMA-3.1-Instruct-8B, Number of random seeds=32026.04 | 52 | |
| Naive Fine-tuningBackbone=LLaMA-3.1-Instruct-8B, Number of random seeds=32026.04 | 51.9 | |
| COMPASS-ECDABackbone=LLaMA-3.1-Instruct-8B, Number of random seeds=32026.04 | 51.9 | |
| Full RetrainingBackbone=LLaMA-3.1-Instruct-8B, Number of random seeds=32026.04 | 51.7 |