As pointed out by the reviewers, these are the strengths and weaknesses of the paper: STRENGTHS The paper proposes algorithms for learning causal trees with intervention data under various assumptions, including infinite observational and interventional data, finite interventional data, allowing K interventions, and limiting the tree nodes that can be intervened on. There is a theoretical analysis on the bounds for the number of required interventions. The paper is overall clearly written. FOR IMPROVEMENT The main concern about this paper is the applicability of the proposed algorithms since they focus only on very specific type of causal graphs (causal trees with no v-structure). The authors should discuss the significance of being able to learn such graphs. Other points that should be discussed are the possibility of having several tree components, possible extensions to multinomials, and comparison with competing methods.