Skip to content
tezvyn:

🤖AI & ML

Artificial intelligence, machine learning, and data science

298 bites

Test yourself: Top 30 intermediate AI & ML interview questionsMultiple choice, with the correct answer and why it is correct on every question. Free, no sign-in.

Intermediate interview questions in AI & ML, page 13

intermediate2 min read

Design a centralized model registry for a large enterprise

Tests ML artifact governance at scale. Strong answers cover immutable versioned artifacts with dependency manifests, a framework-agnostic API, and pluggable deployment targets. Red flag: treating models as opaque files without environment reproducibility.

intermediate2 min read

CLIP's contrastive objective and zero-shot classification

Train image and text encoders to align matched pairs and repel mismatched ones in a shared space; classify zero-shot by comparing an image to text prompts of class names.

intermediate1 min read

Primary metric up, guardrail down: ship or not?

Tie metrics to business value, weigh short-term lift against retention damage, use guardrails and an overall evaluation criterion.

intermediate2 min read

Designing a multi-object tracker

Detect per frame, predict motion with a filter, associate via IoU and appearance, manage track lifecycle, handle occlusion with re-ID.

intermediate2 min read

Why avoid global Python dependencies for ML, and how do containers help?

This probes environment isolation and reproducibility in ML. A strong answer cites global dependency conflicts, system library skew, and brittle environments; then notes containers freeze the full stack for deterministic deployment.

intermediate2 min read

Evaluating faithfulness and compositionality in multimodal models

Use targeted probes with hard negatives, attribute-relation binding tests, and structured grounding checks; note BLEU rewards surface overlap not correctness.

intermediate1 min read

Catalyst Optimizer and Project Tungsten in Spark

Catalyst transforms logical plans with rules, picks physical plans by cost; Tungsten optimizes execution with off-heap memory and codegen.

intermediate1 min read

One-stage vs two-stage detectors

One-stage predicts boxes directly for speed; two-stage proposes then refines for accuracy; focal loss narrows the gap.

intermediate2 min read

How do you configure Docker for host GPU access and CUDA libraries?

This tests GPU passthrough via the NVIDIA Container Toolkit. Strong answers use nvidia/cuda base images matching the host driver, pass GPUs with --gpus all, and avoid installing drivers inside the container.

intermediate2 min read

Hybrid parallelism for large-model training

Split the model itself via tensor or pipeline parallelism so each replica is smaller, shrinking gradient all-reduce; combine with data parallelism in 2D/3D.

intermediate2 min read

Red teaming a generative model

Deliberately probe for harmful outputs across categories, document jailbreaks, and automate with adversarial prompt generators plus classifier-based judging.

intermediate1 min read

Explain false positives and negatives for churn

False positive flags a loyal customer (wasted incentive), false negative misses a leaver (lost customer), tie to threshold choice.

intermediate1 min read

Non-maximum suppression in detection

Detectors emit many overlapping boxes per object; NMS keeps the highest-scoring box and removes others above an IoU threshold.

intermediate1 min read

Flask/Gunicorn vs Triton/TorchServe for serving

Flask is simple and flexible but lacks dynamic batching, GPU scheduling, and multi-model management; Triton/TorchServe add those plus metrics and versioning.

intermediate2 min read

Preventing PII in LLM outputs: curation, fine-tuning, or guardrails

Favor post-processing guardrails as the enforceable last line, backed by data curation; note each layer's tradeoffs and that defense-in-depth is best.

intermediate1 min read

Design an automated A/B test reporting system

Standardized metric definitions, automated stats with confidence intervals and guardrails, segment breakdowns, a clear ship recommendation.

intermediate2 min read

Single vs multi-image super-resolution

SISR hallucinates detail from one image, multi-image fuses sub-pixel shifts; pixel losses blur, perceptual or GAN losses add realism.

intermediate1 min read

Constitutional AI versus standard RLHF

A written principle set guides self-critique and revision, plus AI feedback (RLAIF) replaces human preference labels.

intermediate1 min read

Explain a loan denial with LIME or SHAP

LIME fits a local surrogate, SHAP attributes the prediction across features via Shapley values, both give per-feature contributions.

intermediate2 min read

Triangulating 3D points from two views

Back-project each 2D point into a ray, intersect rays from both cameras using known poses, solve via DLT and refine by minimizing reprojection error.

We are hiring for this. Every open role lists the topics its interview covers, so you can prepare for the real thing rather than guessing.

See open roles