Binary Search based problem (don’t recall specifics)

The source itself says the specifics were not recalled, so it would be misleading to attach one named algorithm. In an interview, request an example, the desired output, input bounds, and whether the task has a sorted sequence or a monotonic feasible-versus-infeasible condition. Then explain the applicable binary-search pattern. For example, to find a minimum feasible value, maintain candidate bounds low and high; test mid; retain mid when feasible and move low past mid when infeasible. The invariant is that a valid answer remains in the interval, and the loop ends when the interval has one candidate. For exact lookup, the comparison rule is different, so do not reuse first-true updates blindly. Verify empty domain, single candidate, no valid answer, and boundaries. Binary search costs O(log n) comparisons on a sorted array or O(log R) predicate calls on a numeric range, with O(1) extra space. Make the clarification explicit rather than guessing from the company report.