Anthropic Online Test Interview Experience for SDE Position
Interview Experience
This post was last edited by Jing666 on 2025-10-3 15:25. Looks like I failed.
Coding Q4, interview question: distributed worker find mode, but I didn't have time to write the follow-up find median, o
Full Details
This post was last edited by Jing666 on 2025-10-3 15:25. Looks like I failed.
Coding Q4, interview question: distributed worker find mode, but I didn't have time to write the follow-up find median, only shared my thoughts. RL Fundamentals: I didn't see this in the interview questions, so I'm contributing it. The following content requires a score higher than 200. You can already view it. Debugging GRPO code. First, there was a Nan error. I found that the given code didn't do softmax before multinomial sampling, directly using logits. Also, when calculating the normalized advantage, stddev didn't add epsilon. After fixing it, a series of questions about my knowledge gaps began, such as: if ratio = model_logprob - old_logprob, can it still be trained? Why clip the ratio? When will the ratio be clipped? What are the effects of clipping? etc. Then they also asked if the ratio in this code is theoretically 1, why? Printing it out showed it wasn't 1, debugging... I'm guessing I'll fail this round. Coding + Design: The following content requires a score higher than 200. You can already view it. Implement a data batcher that can sample weighted data from a given data registry (a data registry and sampling API are provided). The second question is that the sampling API can take an offset argument, requiring the data batcher to produce a batch file/load from a batch file. The third question is, assuming the batch size is not divisible by the sum of weights, how do you sample? Seeking some points to review interview experiences!!
Topics
More from Anthropic
About Anthropic Interview Reports
This question was reported by a candidate who interviewed at Anthropic. LeakCode aggregates interview reports from 10+ sources, including 1Point3Acres, Glassdoor, LeetCode Discuss, Blind, Reddit, Indeed, and Nowcoder. Each report is translated where necessary, deduplicated against existing entries, and tagged by company, role, round type, and reporting date.
Use this question as one calibration data point, not a memorization target. Companies typically rotate their question pools every 2-4 months; the exact wording of a 2024 question may differ from what you encounter today. The underlying pattern, difficulty level, and follow-up depth at Anthropic are the higher-signal extractions to take from this report.
For broader preparation context, the Anthropic interview process typically includes a recruiter screen, one or two technical phone screens, and a 4-5 round on-site loop covering coding, system design (at L4+ levels), and behavioral. Reports tagged on LeakCode show the round-by-round distribution and typical difficulty calibration. To browse questions filtered by round type and seniority, use the company hub linked above.
How To Practice This Type of Question
Solve similar problems on LeetCode under timed conditions (25-35 minutes per medium difficulty). The goal is pattern recognition: recognize the underlying technique (sliding window, two-pointer, BFS, memoized recursion, etc.) within 60-90 seconds of reading. Strong candidates verbalize their hypothesis out loud before coding, then iterate based on feedback. Weak candidates dive into implementation immediately, lose time on the wrong approach, and run out of time for follow-ups.
Companies update their question pools every 2-4 months. The exact wording of any given question may have been retired by the time you interview. Focus your prep on the pattern, not the specific problem. The patterns that appear in Anthropic reports consistently are the ones worth investing in; one-off niche problems are not.
During Your Anthropic Round
Apply the standard interview round template: clarify requirements (2-3 minutes), state your approach out loud and confirm direction with the interviewer (3-5 minutes), code with narration (15-25 minutes), test with concrete examples including edge cases (5 minutes), discuss optimization or trade-offs if time permits (5 minutes). This template is universally accepted across FAANG and adjacent companies; deviating from it produces weaker interviewer feedback signal.
The single most predictive failure mode in Anthropic reports tagged "no hire": not asking clarifying questions. Interviewers are explicitly trained to weight this. Strong candidates ask 3-5 clarifying questions even on problems that look obvious; weak candidates dive into code immediately. The clarifying-question check is often the first signal recorded in the interviewer's written notes.