[2511.21667] Escaping the Verifier: Learning to Reason via Demonstrations
https://arxiv.org/abs/2511.21667
https://arxiv.org/abs/2511.21667
arXiv.org
Escaping the Verifier: Learning to Reason via Demonstrations
Training Large Language Models (LLMs) to reason often relies on Reinforcement Learning (RL) with task-specific verifiers. However, many real-world reasoning-intensive tasks lack verifiers, despite...
reality is really weird as fuck, it is crazy that it is this mathematical, it pretends to be chaotic, but it follows certain rules of uncertainty, the rules themselves are probabilistic, even the randomness has structure, there are a lot of emergent things out of all this and something like love exists, what the fuck
I'm scared of AI in a way which you may find funny, I know it will be helpful to me, but I'll take credit for things it will do
I am realising lately that the real skill is shifting
knowing what to ask, when to push back, and how to judge the output, that's becoming the craft.
I am realising lately that the real skill is shifting
knowing what to ask, when to push back, and how to judge the output, that's becoming the craft.
Introduction — An Introduction to Applied Bioinformatics
https://readiab.org/introduction.html#
https://readiab.org/introduction.html#
Your value increasingly lives in the gradient between what you can specify and what you can verify, not in what you can execute.
[2512.10988] Mathematics of natural intelligence
https://arxiv.org/abs/2512.10988
https://arxiv.org/abs/2512.10988
arXiv.org
Mathematics of natural intelligence
In the process of evolution, the brain has achieved such perfection that artificial intelligence systems do not have and which needs its own mathematics. The concept of cognitome, introduced by...
💡 Remember Box
artificial general cleverness
Intelligence creates invariants. Cleverness exploits correlations.
Frontiers | ADHD as a circadian rhythm disorder: evidence and implications for chronotherapy
https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2025.1697900/full
https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2025.1697900/full
Frontiers
Frontiers | ADHD as a circadian rhythm disorder: evidence and implications for chronotherapy
Accumulating evidence indicates that circadian rhythm dysfunction is a clinically significant and highly prevalent phenotype in a substantial subgroup of ind...
what human activity remains bottlenecked even with AGI, and has fragmented tooling?
Pre-AGI success selected heavily for sustained sequential execution, the ability to grind through boring intermediate steps without dropping the thread.
ADHD brains are notoriously bad at this, but often better at rapid context-switching, pattern-matching across domains, and high-novelty ideation.
ADHD brains are notoriously bad at this, but often better at rapid context-switching, pattern-matching across domains, and high-novelty ideation.
Human intelligence is super-specialized for the physical world, and our feeling of generality is an illusion
We only seem general because we can't imagine the problems we're blind to
We only seem general because we can't imagine the problems we're blind to