April
Papers
- InternLM2 Technical Report
- Towards Self-Assembling Artificial Neural Networks through Neural Developmental Programs
- wav2vec: Unsupervised Pre-training for Speech Recognition
Links
- https://arxiv.org/abs/2312.06681 Steering Llama 2 via Contrastive Activation Addition
- https://www.connectedpapers.com/ Explore connected papers in a visual graph
- https://people.idsia.ch/~juergen/who-invented-backpropagation.html Who Invented Backpropagation?
- https://til.simonwillison.net/llms/embed-paragraphs Embedding paragraphs from my blog with E5-large-v2
- https://mini-gemini.github.io/ Mini-Gemini - Mining the Potential of Multi-modality Vision Language Models
- https://cma-es.github.io/ The CMA Evolution Strategy
- https://github.com/vasturiano/3d-force-graph 3D Force-Directed Graph
- https://pytorch-geometric.readthedocs.io/en/latest/ PyG Documentation (PyTorch Geometric) - Train Graph Neural Networks (GNNs)
- https://www.neuronpedia.org/ Neuronpedia is an open platform for interpretability research
- https://www.openphilanthropy.org/research/reasoning-transparency/ Reasoning Transparency
- https://medium.com/@syoya/what-happens-in-sparse-autencoder-b9a5a69da5c6 What happens in Sparse Autoencoder
- https://www.alignmentforum.org/posts/f9EgfLSurAiqRJySD/open-source-sparse-autoencoders-for-all-residual-stream Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
- https://jonathanbgn.com/2021/09/30/illustrated-wav2vec-2.html An Illustrated Tour of Wav2vec 2.0
- https://cameronrwolfe.substack.com/p/llama-2-from-the-ground-up LLaMA-2 from the Ground Up
- https://lightning.ai/blog/gradient-accumulation/ Finetuning LLMs on a Single GPU Using Gradient Accumulation
- https://medium.com/@anuj_shah/through-the-eyes-of-gabor-filter-17d1fdb3ac97 Through The Eyes of Gabor Filter
- https://remyhax.xyz/posts/bggp4-quantum-rat/ BGGP4: PleaseMom, QUANTUM, Rat?
- https://incidentdatabase.ai/ AI Incident Database
- https://towardsdatascience.com/why-and-how-to-achieve-longer-context-windows-for-llms-5f76f8656ea9 Why and How to Achieve Longer Context Windows for LLMs
- https://rome.baulab.info/ Locating and Editing Factual Associations in GPT
- https://www.lesswrong.com/posts/AcKRB8wDpdaN6v6ru/interpreting-gpt-the-logit-lens Interpreting GPT: the logit lens
- https://www.neelnanda.io/mosaic Induction Mosaic
- https://www.alignmentforum.org/posts/bBuBDJBYHt39Q5zZy/decision-transformer-interpretability#Attention_and_Head_Ablation Decision Transformer Interpretability
- https://towardsdatascience.com/how-to-interpret-gpt2-small-76e0536a588a How to Interpret GPT2-Small
- https://neelnanda-io.github.io/TransformerLens/ Transformer Lens
- https://www.neelnanda.io/mechanistic-interpretability/glossary A Comprehensive Mechanistic Interpretability Explainer & Glossary
- https://www.lesswrong.com/posts/TvrfY4c9eaGLeyDkE/induction-heads-illustrated Induction heads - illustrated
- https://substack.recursal.ai/cp/143397465 Dear VC’s, please stop throwing money at AI founders with no commercial plan, besides AGI
- https://udlbook.github.io/udlbook/ Understanding Deep Learning (book)
- https://www.math.ias.edu/avi/book Mathematics and Computation (book)
- https://clemenswinter.com/2024/04/07/the-simple-beauty-of-xor-floating-point-compression/ The Simple Beauty of XOR Floating Point Compression
- https://paulbupejr.com/developing-the-optigap-sensor-system/ R&D Case Study: Developing the OptiGap Sensor System
- https://eprint.iacr.org/2024/555 Quantum Algorithms for Lattice Problems
- https://www.lesswrong.com/tag/orthogonality-thesis Orthogonality Thesis
- https://en.wikipedia.org/wiki/Unconventional_computing Unconventional computing
- https://web.archive.org/web/20120805130100/singularity.org/files/GISAI.html General Intelligence and Seed AI 2.3
- https://time.is/Anywhere_on_Earth Time Anywhere on Earth
- https://arena3-chapter1-transformer-interp.streamlit.app/ Chapter 1: Transformer Interpretability
- https://transformer-circuits.pub/2023/interpretability-dreams/index.html Interpretability Dreams
- https://10print.org/ 10 PRINT CHR$(205.5+RND(1)); : GOTO 10