danielmamay / grokkingLinks
Implementation of OpenAI's 'Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets' paper.
☆36Updated last year
Alternatives and similar repositories for grokking
Users that are interested in grokking are comparing it to the libraries listed below
Sorting:
- unofficial re-implementation of "Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets"☆78Updated 2 years ago
- Omnigrok: Grokking Beyond Algorithmic Data☆58Updated 2 years ago
- nanoGPT-like codebase for LLM training☆94Updated 3 weeks ago
- Unofficial but Efficient Implementation of "Mamba: Linear-Time Sequence Modeling with Selective State Spaces" in JAX☆81Updated last year
- Language models scale reliably with over-training and on downstream tasks☆97Updated last year
- ☆82Updated last year
- ☆45Updated last year
- Evaluation of neuro-symbolic engines☆35Updated 10 months ago
- ☆29Updated 2 months ago
- ☆67Updated 5 months ago
- ☆26Updated 2 years ago
- The simplest, fastest repository for training/finetuning medium-sized GPTs.☆128Updated 3 weeks ago
- ☆28Updated last year
- ☆66Updated 2 years ago
- Sparse Autoencoder Training Library☆52Updated last month
- Experiments on the impact of depth in transformers and SSMs.☆30Updated 7 months ago
- A MAD laboratory to improve AI architecture designs 🧪☆116Updated 5 months ago
- ☆52Updated last year
- Code for NeurIPS 2024 Spotlight: "Scaling Laws and Compute-Optimal Training Beyond Fixed Training Durations"☆73Updated 7 months ago
- Mixture of A Million Experts☆46Updated 10 months ago
- A centralized place for deep thinking code and experiments☆84Updated last year
- ☆26Updated 2 years ago
- Codes for the paper The emergence of clusters in self-attention dynamics.☆15Updated last year
- Understand and test language model architectures on synthetic tasks.☆197Updated 3 months ago
- ☆26Updated last year
- Attribution-based Parameter Decomposition☆23Updated last week
- ☆37Updated last year
- ☆31Updated 7 months ago
- Revisiting Efficient Training Algorithms For Transformer-based Language Models (NeurIPS 2023)☆80Updated last year
- Code to reproduce "Transformers Can Do Arithmetic with the Right Embeddings", McLeish et al (NeurIPS 2024)☆190Updated last year