Merge intervals test cases
An AI agent wrote a Java method `mergeIntervals(int[][] intervals)` that merges overlapping intervals and returns the merged set. The example in the docstring works. You want to verify it with a small, sharp set of inputs rather than reading the sort-and-merge loop. Which inputs do you choose to expose a likely bug?
Implement
merge_intervals(intervals: list[list[int]]) → list[list[int]]Examples
in
[[[1,3],[2,6],[8,10],[15,18]]]out[[1,6],[8,10],[15,18]]in
[[[1,2],[2,3]]]out[[1,3]]in
[[[1,10],[2,3],[4,5]]]out[[1,10]]What a strong answer looks like
Treat the AI’s output as a draft to verify, not an answer to trust. Name the specific flaw and the input that triggers it, say how you’d catch it (tests, edge cases, reading critically), and how you’d re-prompt or decompose to get it right.
0:00 of about 16 min
Vibe & agentic: describe the solution in plain language (or narrate it) and the coach grades your approach.
Which questions mattered is sealed until you submit. Telling you now would just be handing over the edge cases.
Run or narrate your approach, then ask the coach.