Deschide articolul original într-un tab nou Schimbarea acestei opțiuni se aplică imediat și reîncarcă pagina.
LECTURA
← Înapoi la Urmăritori

Posts

inwaves.io/posts/feed.xml
Vizitează site-ul

Cel mai recent articol

Ultima verificare

Urmăritori 0 urmăritori
Articole: 15

  1. Conceptions of the Heavens

    (Posts)

    Asking whether LLMs can do research leads down at least one rabbit hole: what is research?

  2. fab: how to do (alignment) research at scale

    (Posts)

    Over the last month, I've been working on a project I'm calling fab.1 Nominally, it's an interface that enables a human researcher to make sense of research produced by many agents running in parallel.

  3. Thoughts on 'Why Greatness Cannot Be Planned'

    (Posts)

    Why Greatness Cannot Be Planned is a 2015 book by Kenneth O.

  4. Building a general purpose evaluator

    (Posts)

    Atla Selene Mini: A General Purpose Evaluation Model We introduce Atla Selene Mini, a state-of-the-art small language model-as-a-judge (SLMJ).

  5. On evals

    (Posts)

    Over the last year or so, I’ve been working on evaluating large language models’ capabilities, especially from a safety perspective.

  6. Understanding mixture-of-depths

    (Posts)

    Google Deepmind recently released this paper: Mixture-of-Depths: Dynamically allocating compute in transformer-based language models.

  7. What sorts of systems can be deceptive?

    (Posts)

    The following is a crosspost from the alignment forum.

  8. Why you might expect homogeneous takeoff: evidence from ML research

    (Posts)

    The below is a cross-post from the alignment forum.

  9. Inductive bias of neural networks on 1D regression: an empirical examination

    (Posts)

    The full write-up of my dissertation is now public here.

  10. Non-factorised identifiable variational autoencoders for causal discovery and out-of-distribution generalisation

    (Posts)

    Abstract: "In many situations, given a set of observations, we would like to find the factors which cause or influence the data we observe.

  11. Are graph neural networks (GNNs) fundamentally bottlenecked?

    (Posts)

    Generally speaking, the performance of neural networks on various tasks scales with their depth.

  12. Sharpness-aware minimisation and label noise

    (Posts)

    Do flat minima generalise better?

  13. Comparing GPT-3 and RNNs as probabilistic generative models

    (Posts)

    An investigation of transformers' and RNNs' ability to model long-term dependencies through the lens of probabilistic modelling.

  14. Investigating short-term climate forecasts with surrogate modelling

    (Posts)

    This project uses Gaussian processes (GPs) as surrogate models for accurate climate forecasting.

  15. Intelligent game playing with AI

    (Posts)

    The full write-up of my undergraduate thesis is available here.