Beren Millidge's profile photo

Beren Millidge

Featured in: Favicon plos.org Favicon biorxiv.org

Articles

  • May 12, 2024 | beren.io | Beren Millidge

    Last year, I wrote a quick post investigating the ‘unconditioned’ distribution of LLMs in the OpenAI API, where the ‘unconditioned distribution’ is simply the distribution of LLM outputs following the empty string – or beginning of sequence token. My intuition here was that this gives some idea of what the LLM thinks is ‘intrinsically most likely’ and is likely to be the kind of thing seen most often in training.

  • Mar 16, 2024 | beren.io | Beren Millidge

    After spending a lot of time with language models, I have come to the conclusion that tokenization in general is insane and it is a miracle that language models learn anything at all. To drill down into one specific example of silliness which has been bothering me recently, let’s look at how the GPT2 tokenizer (which is also used for GPT3 as far as I know) tokenizes integers.

  • Mar 3, 2024 | biorxiv.org | Christopher Buckley |Tim Lewens |Michael Levin |Beren Millidge

    AbstractEvolution by natural selection is believed to be the only possible source of spontaneous adaptive organisation in the natural world. This places strict limits on the kinds of systems that can exhibit adaptation spontaneously, i.e. without design. Physical systems can show some properties relevant to adaptation without natural selection or design.

  • Dec 23, 2023 | beren.io | Beren Millidge

    Epistemic status: pretty uncertain. There is a lot of fairly unreliable data in the literature and I make some pretty crude assumptions. Nevertheless, I would be surprised though if my conclusions are more than 1-2 OOMs off though. The brain is currently our sole example of an AGI. Even small animals with tiny brains often exhibit impressive degrees of general intelligence including flexibility and agentic behaviour in a complex and uncertain world.

  • Oct 4, 2023 | beren.io | Beren Millidge

    Minor terminological nitpick. People often describe the LLM as ‘hallucinating’ information whenever it makes up information which seems like it should fit for a given query even when it is trivially false. While evocative, this isn’t actually correct terminology. We already have a perfectly word for this precisely phenomenon in psychology: confabulation.

Contact details

Socials & Sites

Try JournoFinder For Free

Search and contact over 1M+ journalist profiles, browse 100M+ articles, and unlock powerful PR tools.

Start Your 7-Day Free Trial →