← Citadel Interview Insights

Citadel·Software Engineer·Technical Phone Screen·Senior

Senior
Apr 2026

Summary

Citadel research engineer interview touched on probability theory with some genuinely tricky conceptual angles. The conditional probability framing was the part that stuck with me most.

Questions Asked (1)

Q1

How can you use conditional probability to reframe a uniform distribution on (0, a) in terms of a standard uniform distribution on (0, 1)?

Algorithms & Data StructuresTechnical Trade-offs
Author's notes

The key insight is that U(0,a) is just U(0,1) conditioned on being less than a, which is a clean way to think about it but not where my brain went first.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining the problem: you have a random variable X ~ Uniform(0, a) and want to express it in terms of U ~ Uniform(0, 1). Use the probability integral transform or a scaling argument: show that X = a * U, and derive the conditional distribution of X given U. Then explain how conditioning on U allows you to reframe the uniform distribution on (0, a) as a scaled version of the standard uniform.

Pro tip: Emphasize that this is a fundamental technique in inverse transform sampling and is widely used in simulations and probabilistic modeling. Mention that understanding this relationship is crucial for generating random variables from arbitrary distributions.

1. Define the random variables

Let U ~ Uniform(0, 1) and X ~ Uniform(0, a). Clearly state their probability density functions: f_U(u) = 1 for 0 < u < 1, and f_X(x) = 1/a for 0 < x < a.

2. Establish the transformation

Consider the transformation X = a * U. Show that this maps U from (0,1) to X in (0,a). Verify that the CDF of X matches that of a uniform on (0,a): P(X ≤ x) = P(aU ≤ x) = P(U ≤ x/a) = x/a for 0 < x < a.

3. Use conditional probability to reframe

Express the conditional distribution of X given U = u. Since X = aU, given U = u, X is deterministically a*u. Thus, P(X ≤ x | U = u) = 1 if a*u ≤ x, else 0. This shows that conditioning on U gives a degenerate distribution at a*u, but the marginal distribution of X is Uniform(0,a).

4. Generalize to inverse transform sampling

Explain that this is a special case of the probability integral transform: if U ~ Uniform(0,1) and F is a CDF, then X = F^{-1}(U) has distribution F. Here, F(x) = x/a for x in (0,a), so F^{-1}(u) = a*u.

5. Discuss implications and applications

Highlight that this reframing allows generating samples from Uniform(0,a) using a standard uniform generator. It also illustrates how conditional probability can simplify understanding of scaling and transformations.

Key Points to Mention

  • Probability integral transform: If U ~ Uniform(0,1), then X = F^{-1}(U) has CDF F.
  • Scaling property: If U ~ Uniform(0,1), then aU ~ Uniform(0,a).
  • Conditional probability: P(X ≤ x | U = u) = I(a*u ≤ x), showing deterministic relationship.
  • Inverse transform sampling: A method to generate random variables from any distribution given its inverse CDF.
  • Uniform distribution properties: constant density, linear CDF.
  • Application in simulations and Monte Carlo methods.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.