rhubarbwu / linguistic-collapse
Codebase for arXiv:2405.17767, based on GPT-Neo and TinyStories.
☆11Updated 4 months ago
Alternatives and similar repositories for linguistic-collapse:
Users that are interested in linguistic-collapse are comparing it to the libraries listed below
- ☆90Updated last month
- ☆23Updated last month
- ☆66Updated 3 years ago
- Source code of "Task arithmetic in the tangent space: Improved editing of pre-trained models".☆98Updated last year
- DataInf: Efficiently Estimating Data Influence in LoRA-tuned LLMs and Diffusion Models (ICLR 2024)☆63Updated 5 months ago
- Code for my NeurIPS 2024 ATTRIB paper titled "Attribution Patching Outperforms Automated Circuit Discovery"☆30Updated 9 months ago
- AI Logging for Interpretability and Explainability🔬☆108Updated 9 months ago
- LLM finetuning in resource-constrained environments.☆47Updated 9 months ago
- A library for efficient patching and automatic circuit discovery.☆59Updated last month
- NeuroSurgeon is a package that enables researchers to uncover and manipulate subnetworks within models in Huggingface Transformers☆40Updated last month
- Influence Functions with (Eigenvalue-corrected) Kronecker-Factored Approximate Curvature☆140Updated 8 months ago
- ☆28Updated 8 months ago
- ☆213Updated 5 months ago
- ☆78Updated last week
- A Mechanistic Understanding of Alignment Algorithms: A Case Study on DPO and Toxicity.☆71Updated 3 weeks ago
- ☆197Updated last year
- Official repository for our paper, Transformers Learn Higher-Order Optimization Methods for In-Context Learning: A Study with Linear Mode…☆15Updated 4 months ago
- ☆17Updated 3 weeks ago
- ☆47Updated last year
- ☆30Updated 3 months ago
- ☆37Updated last year
- ☆93Updated last year
- Sparse probing paper full code.☆55Updated last year
- This is the repository for "Model Merging by Uncertainty-Based Gradient Matching", ICLR 2024.☆27Updated 10 months ago
- Efficient empirical NTKs in PyTorch☆18Updated 2 years ago
- ☆66Updated 4 months ago
- ☆217Updated 10 months ago
- ☆78Updated last year
- Röttger et al. (NAACL 2024): "XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models"☆95Updated last month
- Implementation of PaCE: Parsimonious Concept Engineering for Large Language Models (NeurIPS 2024)☆35Updated 4 months ago