Curious cases of financial engineering in biotech
IntroductionFor $250 million and ten years of your life, you may purchase a lottery ticket.…
Category: Opinion
From nothing to important actions: agents that act morally
You may start reading here, or jump to the “Comment” section or to the “Takeaways”.…
The other paper that killed deep learning theory
Yesterday, I wrote about the state of deep learning theory circa 2016,[1] as well as the…
Emergent misalignment evident in activations at low poisoning doses – long before behavioral checks flag it
Status: Exploration with adapters for Llama 3.2-3B + 3.1-8B at various poisoning dose levels using…
Massapequa ACX Meetup
Saturday, May 9, at 7 pm. At my house, 47 Clinton Pl., Massapequa, NY 11758.…
Retrospective on my unsupervised elicitation challenge
This post contains spoilers for the unsupervised elicitation challenge of getting Claude to get my…
Alignment Faking Replication and Chain-of-Thought Monitoring Extensions
In this post, I present a replication and extension of the alignment faking model organism…
Training a Transformer to Compose One Step Per Layer (and Proving It)
I'm working on an experiment comparing the internal representations of two architectures when solving a…
AI for life strategy advice: a personal experiment
If AI could provide good decision advice and life planning guidance more broadly, this would…
Spontaneous introspection in output tampering
Epistemic status: This is a side project which I spent a relatively small effort on.…