I talk about the enormous potential of generative models to improve downstream predictive tasks in this excellent article by @anilananth
RESEARCH
-
AI-generated content detection methods easily circumvented
By
–
Me reitero en lo dicho, cualquier opción de detectar contenido generado por IA actualmente puede ser fácilmente contratado.
-
Training People-Diffusion Model for Image Generation
By
–
Progress on training my own model called people-diffusion: 1) New Year celebrations in China
2) Protests in Brazil
3) Indigenous Bolivian
4) Woman at Times Square -
Invisible Watermarking Systems Easily Defeated by Adversarial Methods
By
–
Por ejemplo, me estáis compartiendo un vídeo donde se propone marcar los textos de ChatGPT de forma invisible para poder detectar que es artificial. El problema, como comenta aquí Jeremy, es que estos sistemas una vez creados, son muy fáciles de superar.
-
Can AI-Generated Content Be Reliably Detected?
By
–
¿Cómo podemos detectar si una imagen o un texto ha sido creado por una IA? ¿Se puede usar otra IA? Esto es algo que muchos me preguntáis y malas noticias: NO SE PUEDE Al menos por ahora no hay método satisfactorio que te permita identificar si algo es artificial… [1/n]
-
Dutch Tech Leaders Showcase Responsible AI at CES 2023
By
–
The Dutch are ready for #CES2023. Proud to be part of the #NLatCES. Technology has the potential to tackle some of society’s biggest challenges. Meet us at Hall A-C 55332 and see how we use #responsibletech to #acceleratethechange. #AI #ML #computervision
-
Anthropic Hiring Research Engineers and Scientists for Interpretability
By
–
We’re also actively hiring research engineers/scientists to work with us on interpretability. If you’re interested, we’d encourage you to apply!
Research engineer: https://
jobs.lever.co/Anthropic/436c
a148-6440-460f-b2a2-3334d9b142a5
…
Research scientist: https://
jobs.lever.co/Anthropic/eb9e
6d83-626c-4f59-8a0e-fa7c413b2014
… -
Thanks to Adam S Jermyn for reproducing and extending results
By
–
Thanks to @AdamSJermyn for his comments reproducing and extending these results! https://
transformer-circuits.pub/2023/toy-doubl
e-descent/index.html#comment-jermyn-1
… -
Mechanistic Theory of Memorization: Open Questions and Research Directions
By
–
We hope these results are a step towards a mechanistic theory of memorization. There are many open questions, such as understanding the loss spike, or what happens when only a subset of the data is repeated.
-
Model Capacity and Double-Descent: Strategy Transitions in ML
By
–
Models struggle to transition between these strategies, as exhibited by a spike in test loss. This spike moves to larger datasets as one increases model capacity. This is a clear signature of double-descent, a phenomenon that is now well-known in the ML literature.