Top postsTop post
TLDR: InsanityBench is a benchmark of handcrafted cryptic puzzles trying to measure creative conceptual leaps as required in, for example, scientific breakthroughs - SOTA models score around 10% - the benchmark is very much in its earliest form and will need to be properly scaled to reduce variance. You are...
Yesterday I attended a talk by Franziska Boenisch on training data memorization in diffusion models. The short version: diffusion models memorize a small percentage of their training data verbatim, reproducible with specific prompts regardless of noise seed. Privacy concern, etc. I was mostly interested in the adversarial case - say...
Imagine a friend gets kidnapped by mobsters who are surprisingly obsessed with human psychology. They phone you with a deal: your friend will survive if and only if you show up and play a game. The game is simple: a random number between 0 and 100 is generated, and if...
Main thesis: Discrete token vocabularies don't lose information so much as they allow information to be retained in the first place. By removing minor noise and singling out major noise, errors become identifiable and therefore correctable, which continuous latent representations fundamentally cannot offer. The Bandwidth Intuition (And Why It's Incomplete)...
An on-policy, sample-efficient NLP post-training approach not requiring verification The current state I once tried teaching a simulated 3D humanoid how to dunk[1]using RL - truly the most effective use of my time. If I took a single thing away from that, it was that designing the best reward function...