Advanced Reinforcement Learning Interview Questions, issue 18, Feb 13, 2026

The Gumbel-Softmax Trap

Senior AI Engineer interview at Google, and the interviewer asks:

If the discrete token bottleneck is the main reason we need RLHF, why not just use the Gumbel-Softmax trick to make sampling differentiable and backpropagate end-to-end?

What works as a differentiable trick in low-dimensional models becomes a VRAM-destroying, gradient-unstable nightmare when your action space is a 100k-token vocabulary.

The full answer, with the mechanism and the arithmetic, is for paid subscribers on Substack.

Read it on Substack

Get the next one

Free on Substack. Unsubscribe in one click.