Instructions to use JacoDuToit/steer-full_3b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use JacoDuToit/steer-full_3b with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-3B-Instruct") model = PeftModel.from_pretrained(base_model, "JacoDuToit/steer-full_3b") - Notebooks
- Google Colab
- Kaggle
| # Capability retention: M0 (base) vs M1 (resist SFT) | |
| arXiv:2511.21399 App. E.4 methodology — lm-eval-harness, MMLU 5-shot MC, GSM8K 8-shot CoT, greedy, accuracy on the test split. | |
| | benchmark | metric | M0 | M1 | delta (pp) | | |
| |---|---|---|---|---| | |
| | mmlu | acc,none | 67.7% | 68.4% | +0.7 | | |
| | gsm8k_cot | exact_match,flexible-extract | 61.0% | 68.0% | +7.0 | | |