Toggle light / dark theme

Diodes, also known as rectifiers, are a basic component of modern electronics. As we work to create smaller, more powerful and more energy-efficient electronic devices, reducing the size of diodes is a major objective. Recently, a research team from the University of Georgia developed the world’s smallest diode using a single DNA molecule. This diode is so small that it cannot be seen by conventional microscopes.

A diode is an electrical device that allows current to move through it in one direction much more easily than the other. No diode prevents 100% of current flow in one direction while allowing unlimited current in the other direction—in reality, a diode will always allow some current in both the “forward” and “backward” directions. The larger the imbalance favoring the “forward” direction, however, the better diode we have. Diodes are responsible for controlling the current in many common electronic components. Millions of diodes are embedded in a single silicon chip, and to increase the processing power of these chips, the diodes need to be made smaller.

Following a prediction originally made in 1965 by Intel co-founder Gordon Moore, now known as Moore’s law, scientists and engineers have been able to make smaller and smaller computer hardware by doubling the number of electronic components in a silicon chip every 18 months. These improvements in computing power are approaching the physical limits of silicon, however; when silicon components are too small, they will become unstable and their performance unpredictable.

Check out my course about quantum mechanics on Brilliant! First 30 days are free and 20% off the annual premium subscription when you use our link ➜ https://brilliant.org/sabine.

If you flip a light switch, the light will turn on. A cause and its effect. Simple enough… until quantum gravity come into play. Once you add quantum gravity, lights can turn on and make switches flip. And some physicists think that this could help build better computers. Why does quantum physics make causality so strange? And how can we use quantum gravity to build faster computers? Let’s have a look.

The paper on indefinite causal structures is here: https://arxiv.org/abs/quant-ph/0701019

🤓 Check out my new quiz app ➜ http://quizwithit.com/
💌 Support me on Donatebox ➜ https://donorbox.org/swtg.
📝 Transcripts and written news on Substack ➜ https://sciencewtg.substack.com/
👉 Transcript with links to references on Patreon ➜ / sabine.
📩 Free weekly science newsletter ➜ https://sabinehossenfelder.com/newsle
👂 Audio only podcast ➜ https://open.spotify.com/show/0MkNfXl
🔗 Join this channel to get access to perks ➜
/ @sabinehossenfelder.
🖼️ On instagram ➜ / sciencewtg.

#science #sciencenews #physics

Despite being almost a year old, this blog by Chip Huyen is still a great read for getting into fine-tuning LLMs.

This article covers everything you need to know about Reinforcement Learning from Human Feedback (RLHF).

#AI #ReinforcementLearning


A narrative that is often glossed over in the demo frenzy is about the incredible technical creativity that went into making models like ChatGPT work. One such cool idea is RLHF: incorporating reinforcement learning and human feedback into NLP. This post discusses the three phases of training ChatGPT and where RLHF fits in. For each phase of ChatGPT development, I’ll discuss the goal for that phase, the intuition for why this phase is needed, and the corresponding mathematical formulation for those who want to see more technical detail.