Skip to content

Tracks Drills Topics Technologies Tech stacks Skill areas

Tracks Drills Topics

Loading

Aporeon

Technical interview practice, from one developer to another.

About Editorial standards Support Privacy Terms Refund policy

AI Fundamentals

Models, tokens, context, sampling, embeddings, limitations, model selection, and dependable application boundaries.

20 drillsPractice one at a timeOrdered pathConcept questionComparison questionArchitecture scenarioDebugging scenario

Recommended start

What an LLM predicts

Defines a large language model through next-token prediction and connects that mechanism to useful language behavior and its limits.

Easy~5 minStep 1 of 20

Practice this drill Choose another

Search

20 of 20 shown

1What an LLM predictsRecommended startDefines a large language model through next-token prediction and connects that mechanism to useful language behavior and its limits.Concept questionEasy~5 minArtificial IntelligenceGenerative AILarge Language ModelsPractice

Practice order

20 drillsOne at a time

1What an LLM predictsUp next / Easy / ~5 min
2Tokens and tokenizationEasy / ~5 min
3The context windowEasy / ~5 min

2

Tokens and tokenizationExplains tokens as model-specific text units and connects tokenization to context limits, output limits, latency, cost, and text behavior.Concept questionEasy~5 minArtificial IntelligenceLarge Language Models

Practice

3The context windowDescribes the context window as the bounded request state containing instructions, conversation, retrieved evidence, tool results, and output capacity.Concept questionEasy~5 minArtificial IntelligenceLarge Language ModelsPractice

4Why outputs varyConnects probabilistic decoding and changing service conditions to output variation, reproducibility limits, and appropriate testing.Concept questionMedium~7 minArtificial IntelligenceLarge Language ModelsPractice

5Temperature and samplingExplains how decoding settings reshape token selection and why they must be tuned against task-specific evaluation rather than folklore.Concept questionMedium~7 minArtificial IntelligenceLarge Language ModelsPractice

6EmbeddingsDefines embeddings as model-produced vectors used for similarity and contrasts representation with text generation.Comparison questionMedium~7 minArtificial IntelligenceLarge Language ModelsVector DatabasesPractice

7Ways to adapt an AI featureSeparates runtime model use from model training, fine-tuning, prompt design, and retrieval-augmented generation as different engineering levers.Comparison questionMedium~10 minArtificial IntelligenceGenerative AILarge Language ModelsPractice

8HallucinationDefines hallucination as unsupported or incorrect model output and connects it to evidence, task design, evaluation, and user-visible uncertainty.Concept questionEasy~5 minArtificial IntelligenceGenerative AILarge Language ModelsPractice

9When not to use an LLMIdentifies tasks better served by deterministic code, search, rules, or conventional models because of correctness, latency, cost, or control.Comparison questionMedium~7 minArtificial IntelligenceLarge Language ModelsPractice

10Choose a modelChooses a model from task-specific evaluation, operating constraints, safety, latency, cost, context, modality, and provider requirements.Comparison questionMedium~10 minArtificial IntelligenceGenerative AILarge Language ModelsPractice

11Hosted vs self-hosted modelsCompares managed model APIs with operating open-weight models across control, expertise, privacy, scale, latency, cost, and portability.Comparison questionHard~12 minArtificial IntelligenceGenerative AILarge Language ModelsPractice

12Quality, latency, and costBalances AI quality, tail latency, and cost through explicit product thresholds, model choice, context control, caching, routing, and measurement.Comparison questionHard~12 minArtificial IntelligenceGenerative AILarge Language ModelsPractice

13Use a smaller modelUses task bounds and evaluation evidence to choose smaller models for lower latency, cost, capacity, privacy, or local execution.Comparison questionMedium~7 minArtificial IntelligenceGenerative AILarge Language ModelsPractice

14Capability vs reliabilitySeparates what a model can perform from whether the complete AI feature behaves dependably under real inputs, failures, and constraints.Comparison questionMedium~10 minArtificial IntelligenceLarge Language ModelsPractice

15Deterministic AI boundariesPlaces validation, authorization, limits, state transitions, and approval outside the model while acknowledging that guardrails reduce rather than eliminate risk.Architecture scenarioMedium~10 minArtificial IntelligenceLarge Language ModelsPractice

16Context and output limitsHandles token-budget overflow and incomplete generation through explicit prioritization, rejection, retrieval, summarization, continuation, and validation.Debugging scenarioMedium~10 minArtificial IntelligenceLarge Language ModelsPractice

17Handle model failuresSeparates model and provider outcomes, applies bounded recovery, validates outputs, and exposes useful status and telemetry.Architecture scenarioHard~12 minArtificial IntelligenceLarge Language ModelsPractice

18Roll out a model changeTreats model changes as controlled releases using versioned configuration, fixed evaluations, shadow or canary traffic, monitoring, and rollback.Architecture scenarioHard~12 minArtificial IntelligenceLarge Language ModelsPractice

19Diagnose AI latency and costDiagnoses an AI regression by decomposing request stages, configuration changes, token growth, retries, routing, tools, retrieval, caching, and provider behavior.Debugging scenarioHard~15 minArtificial IntelligenceLarge Language ModelsPractice

20Design a first LLM featureDesigns a bounded and reversible first AI feature with measurable value, controlled data, validation, evaluation, observability, fallback, and human oversight.Architecture scenarioHard~15 minArtificial IntelligenceGenerative AILarge Language ModelsPractice

4

Why outputs vary

Medium / ~7 min

5Temperature and samplingMedium / ~7 min

6EmbeddingsMedium / ~7 min

7Ways to adapt an AI featureMedium / ~10 min

8HallucinationEasy / ~5 min

+12 more drills in this section

AI Fundamentals - AI Engineering | Aporeon