Search

Few direct matches — filled in with the latest updates.

Tag: #adapters2 results

Adapters Unlock Reliable Self-Interpretation

Adapters Unlock Reliable Self-Interpretation

Lightweight adapters trained on interpretability artifacts enable reliable self-interpretation in frozen LMs. A simple scalar affine adapter outperforms baselines in feature labeling, topic identification, and implicit reasoning decoding. Gains scale with model size, driven mostly by learned bias.

ArXiv AIResearchFeb 12#research#self-interpretation#v1
1% Params Beat Full Fine-Tuning

1% Params Beat Full Fine-Tuning

CoLin introduces a 1% parameter low-rank complex adapter for vision foundation models. It resolves convergence issues in composite matrices with tailored loss. Surpasses full fine-tuning and delta-tuning on detection, segmentation, and classification.

ArXiv AIResearchFeb 12#research#arxiv-ai#v1