How Can Post-Trained Models Continually Learn with Off-Policy Training?
How can post-trained models learn from off-policy data while retaining their existing capabilities?
Don't (Just) Train For Performance
tl;dr: Extended pre-training improves base model performance but can hurt adaptability—the ability to fine-tune effectively while retaining capabilities. This calls for research explore methods to explicitly pre-train for adaptability.
Weights could be helpful for interpretation
tl;dr: We demonstrate how analyzing weight differences between pre- and post-fine-tuned models can detect backdoors and suspicious behavior without needing trigger examples.