Top 29
Prep plan
Updated weekly · Last refresh Oct 17

Meta Research Scientist Interview Questions

The questions to prepare for a Meta Research Scientist interview. Questions from real interview reports rank first. Updated daily.

29questions
~5htotal time
Track your progressSign up free to work through all 29 questions and resume where you left off.
Start practicing free →
1
CodingStart here. 6 questions · ~63 min
Kth Largest Feed ScoreEasy
Practice
Recently asked

Find the kth largest value in an unsorted array using Quickselect for expected O(n) time.

ArraysSortingHeapMeta
Find Pair Indices for TargetEasy
Practice
Recently asked

Find two indices in an unsorted array whose values sum to a target using a hash table in O(n) time.

Hash TablesArraysSearchingMeta
More Coding questions with a free account
2
Machine Learning8 questions · ~84 min
Analyze Self-Attention for Feed RankingHard
Recently asked

Explain and implement scaled dot-product self-attention for Facebook Feed text ranking, including tensor shapes, training setup, and complexity analysis.

Neural NetworksDeep LearningGradient DescentMeta
Stabilize Deep Ranking TrainingMedium
Recently asked

Train a deep neural network for Meta Feed ranking and mitigate vanishing gradients while preserving offline AUC and serving efficiency.

Neural NetworksRegularizationDeep LearningMeta
More Machine Learning questions with a free account

Sign up to see every question

Create a free account to unlock this list and practice real interview questions.

Get my prep plan
3
System Design4 questions · ~42 min
Facebook Page Follow RecommenderHard
Recently asked

Design Facebook's end-to-end recommender for predicting and ranking Pages a person is likely to follow.

ML RankingCold StartdesignMeta
Predict Onsite Event AttendanceHard
Recently asked

Tests system design skills for building a prediction service with data, modeling, and operational considerations.

Meta
More System Design questions with a free account
4
Behavioral & Leadership9 questions · ~94 min
More Behavioral & Leadership questions with a free account
5
More topics2 questions · ~21 min
LLM Post-Training Issue AnalysisHard
Recently asked

Analyze post-training failures such as regression, overfitting, hallucination, instruction loss, and unsafe behavior, then propose an eval-first remedy.

HallucinationLLM EvaluationFine-TuningMeta
Survey Weights ConstructionHard
Recently asked

Construct survey weights using selection probabilities, nonresponse adjustment, calibration, and trimming.

SamplingBiasData AnalysisMeta
The finish line: interview-readyComplete all 29 questions to finish this plan.