🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Hello world! You can read our launch announcement here: anthropic.com/news/announcem…

Our first AI alignment paper, focused on simple baselines and investigations: A General Language Assistant as a Laboratory for Alignment arxiv.org/abs/2112.00861

Our first interpretability paper explores a mathematical framework for trying to reverse engineer transformer language models: A Mathematical Framework for Transformer Circuits: transformer-circuits.pub/2021/framework…

Our first societal impacts paper explores the technical traits of large generative models and the motivations and challenges people face in building and deploying them: arxiv.org/abs/2202.07785

In our second interpretability paper, we revisit “induction heads”. In 2+ layer transformers these pattern-completion heads form exactly when in-context learning abruptly improves. Are they responsible for most in-context learning in large transformers? transformer-circuits.pub/2022/in-contex…

On the @FLIxrisk podcast, we discuss AI research, AI safety, and what it was like starting Anthropic during COVID. futureoflife.org/2022/03/04/dan…

We've trained a natural language assistant to be more helpful and harmless by using reinforcement learning with human feedback (RLHF). arxiv.org/abs/2204.05862

Glad Quanta Magazine highlights progress on induction heads/rigorous interpretability by Chris Olah, Catherine Olsson, @nelhage and others Anthropic. More to come! quantamagazine.org/researchers-gl…

In a new paper, we show that repeating only a small fraction of the data used to train a language model (albeit many times) can damage performance significantly, and we observe a "double descent" phenomenon associated with this. arxiv.org/abs/2205.10487

Transformer MLP neurons are challenging to understand. We find that using a different activation function (Softmax Linear Units or SoLU) increases the fraction of neurons that appear to respond to understandable features without any performance penalty. transformer-circuits.pub/2022/solu/inde…

In "Language Models (Mostly) Know What They Know", we show that language models can evaluate whether what they say is true, and predict ahead of time whether they'll be able to answer questions correctly. arxiv.org/abs/2207.05221

Neural networks often pack many unrelated concepts into a single neuron – a puzzling phenomenon known as 'polysemanticity' which makes interpretability much more challenging. In our latest work, we build toy models where the origins of polysemanticity can be fully understood.

Introducing Claude 2! Our latest model has improved performance in coding, math and reasoning. It can produce longer responses, and is available in a new public-facing beta website at claude.ai in the US and UK.