Anthropic interview questions

Showing 211–225 of 225 results

• Schedule Large and Small Model Inference on an Eight-GPU PoolHard

  Design an inference scheduler for a shared 8-GPU node that co-locates single-GPU small models and all-reduce 8-GPU large models. The design must address gang scheduling, prefix cache reuse, deadline-aware priorities, and overload mitigation.

  AnthropicMachine Learning EngineerML System DesignTechnical Screen
• Schedule instructions on a VLIW pipelineHard

  This question evaluates proficiency in VLIW instruction scheduling, pipeline hazard analysis (RAW/WAR/WAW), register-pressure management, and software-pipelining for loop kernels within the System Design domain of computer architecture and compiler backends.

  AnthropicSoftware EngineerSystem DesignOnsite
• Simulate Threshold Infection EfficientlyHard

  Simulate a synchronous threshold-based infection process on a grid using an incremental multi-source BFS to achieve optimal linear time complexity.

  AnthropicSoftware EngineerCoding & AlgorithmsOnsite
• Simulate stack traces from logsMedium

  This question evaluates stack-based simulation for computing exclusive execution time in nested function call logs and stream-processing concepts for handling equal timestamps, slight reordering, and detection of N consecutive categorical events.

  AnthropicSoftware EngineerCoding & AlgorithmsOnsite
• Solve programming task with follow-upsMedium

  This question evaluates algorithmic problem-solving skills, including core algorithm implementation, handling a consecutive-N-element constraint, and designing an optimization that leverages suffix-related information.

  AnthropicSoftware EngineerCoding & AlgorithmsTechnical Screen
• System Design: Pushing a Multi-Terabyte Snapshot to Thousands of NodesMedium

  Figure out the theoretical floor on how fast a large search-index snapshot can reach every node in a fleet given each machine's shared upload/download bandwidth, then design a chunked, peer-assisted distribution scheme that approaches that floor and copes with stragglers and failures. Tests bandwidth reasoning and evolving a design from a naive baseline.

  AnthropicSoftware EngineerSystem DesignOnsite
• Tell Me About a Metric You Built That Changed a Real DecisionMedium

  A behavioral prompt asking the candidate to walk through a time they invented or redefined a metric, prove it was measuring the right thing, and show it actually changed what the business did next.

  AnthropicData ScientistBehavioral & LeadershipHR Screen
• Time-Based Key-Value StoreMedium

  Design a versioned key-value store that supports storing values at specific timestamps and retrieving the most recent value at or before a given timestamp.

  AnthropicSoftware EngineerCoding & AlgorithmsOnsite
• Walk Through Your Work Background and Explain Your Concerns About AI SafetyMedium

  A recruiter screen question for a machine learning engineering role at an AI research company: give a concise overview of your work background, then explain the concerns you have about AI safety. It tests whether you can tell a focused career story and discuss safety risks with specific mechanisms, calibrated views and a link to the role.

  AnthropicMachine Learning EngineerBehavioral & LeadershipHR Screen
• Walk through a recent technical projectHard

  This question evaluates a candidate's engineering depth and leadership competency by requiring a structured walkthrough of a recent technical project covering problem framing, architectural trade-offs, data modeling and APIs, scalability and reliability, testing and rollout, and measurement of outcomes.

  AnthropicSoftware EngineerBehavioral & LeadershipOnsite
• Why Anthropic and its values?Medium

  This question evaluates a candidate's culture fit, value alignment, integrity under pressure, and critical thinking within the Behavioral & Leadership category for a software engineering role.

  AnthropicSoftware EngineerBehavioral & LeadershipTechnical Screen
• Why Anthropic, What AI Safety Means to You, and AI Misuse: Recruiter Screen PrepMedium

  Prepare for a recruiter screen at an AI safety lab that spends most of its time on motivation and safety views: why you are leaving, why this lab over others, what AI safety means to you, misuse scenarios, the biggest risks and benefits of advanced AI, and an underrated problem. It tests depth, honesty and preparation.

  AnthropicSoftware EngineerBehavioral & LeadershipOnsite
• Why Join an AI Safety-Focused Lab, and What Is Your View on AI Safety?Hard

  A recruiter-call question for a senior software engineer at an AI lab: explain why you want to join this particular company and share your own view on AI safety. It tests whether your motivation is specific and sincere, whether you can state a reasoned, calibrated position on AI risk, and how you connect safety to engineering work.

  AnthropicSoftware EngineerBehavioral & LeadershipOnsite
• Why Work on AI SafetyMedium

  Assesses your motivation and understanding of AI safety, focusing on how engineering skills intersect with frontier risks, mitigation strategies, and intellectual honesty.

  AnthropicSoftware EngineerBehavioral & LeadershipHR Screen
• Word Search in a Letter GridMedium

  This problem tests the ability to implement a backtracking search on a 2D grid, optimized with a Trie data structure to efficiently search for multiple strings simultaneously.

  AnthropicSoftware EngineerCoding & AlgorithmsOnsite