mcleish7 / arithmeticLinks
Code to reproduce "Transformers Can Do Arithmetic with the Right Embeddings", McLeish et al (NeurIPS 2024)
☆189Updated last year
Alternatives and similar repositories for arithmetic
Users that are interested in arithmetic are comparing it to the libraries listed below
Sorting:
- The simplest, fastest repository for training/finetuning medium-sized GPTs.☆126Updated 3 weeks ago
- Understand and test language model architectures on synthetic tasks.☆195Updated 2 months ago
- A MAD laboratory to improve AI architecture designs 🧪☆116Updated 5 months ago
- Code for NeurIPS'24 paper 'Grokked Transformers are Implicit Reasoners: A Mechanistic Journey to the Edge of Generalization'☆193Updated 5 months ago
- ☆130Updated 2 months ago
- Delphi was the home of a temple to Phoebus Apollo, which famously had the inscription, 'Know Thyself.' This library lets language models …☆180Updated this week
- Code for exploring Based models from "Simple linear attention language models balance the recall-throughput tradeoff"☆233Updated 3 months ago
- Language models scale reliably with over-training and on downstream tasks☆97Updated last year
- Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT training☆127Updated last year
- ☆190Updated this week
- Implementation of 🥥 Coconut, Chain of Continuous Thought, in Pytorch☆168Updated 5 months ago
- Repository for the paper Stream of Search: Learning to Search in Language☆146Updated 3 months ago
- ☆174Updated last month
- ☆179Updated last year
- Code for reproducing our paper "Not All Language Model Features Are Linear"☆74Updated 6 months ago
- Tree Attention: Topology-aware Decoding for Long-Context Attention on GPU clusters☆126Updated 5 months ago
- EvaByte: Efficient Byte-level Language Models at Scale☆97Updated last month
- Replicating O1 inference-time scaling laws☆87Updated 5 months ago
- Extract full next-token probabilities via language model APIs☆248Updated last year
- ☆81Updated last year
- ☆95Updated 4 months ago
- ☆166Updated last year
- Create feature-centric and prompt-centric visualizations for sparse autoencoders (like those from Anthropic's published research).☆200Updated 5 months ago
- OpenCoconut implements a latent reasoning paradigm where we generate thoughts before decoding.☆171Updated 4 months ago
- ☆79Updated 9 months ago
- Code for the paper "VinePPO: Unlocking RL Potential For LLM Reasoning Through Refined Credit Assignment"☆159Updated this week
- Pytorch implementation of the PEER block from the paper, Mixture of A Million Experts, by Xu Owen He at Deepmind☆124Updated 9 months ago
- ☆121Updated last year
- Functional Benchmarks and the Reasoning Gap☆86Updated 7 months ago
- Open source interpretability artefacts for R1.☆138Updated last month