Skip to content
ml-recall

#softmax

38 questions from Multi-Agent Systems, Optimization & Calculus, Logistic Regression & Softmax, Neural Network Basics, Attention & Transformers, Text Representation & Embeddings, Language Modeling, PyTorch

Timed test

20 questions · 20 min · score and explanations at the end

Longer test or by difficulty: mixed test with #softmax

Take a test
Q 1 / 38Score 0 / 0

Question 1: Single choice

mediumAI-checkedMulti-Agent Systems

A learner keeps cumulative utilities g∈Rng \in \mathbb{R}^n for its nn actions and plays x=arg max⁡x∈Δn[⟨g,x⟩−ψ(x)]x = \operatorname*{arg\,max}_{x \in \Delta_n}\big[\langle g, x \rangle - \psi(x)\big], with the negative entropy ψ(x)=∑ixilog⁡xi\psi(x) = \sum_i x_i \log x_i as regularizer and Δn\Delta_n the probability simplex. What is xx?

Choose one option