General comment:
This paper explores what factors lead to children acting compassionately. It is great the studies are pre-registered, and that there are many studies asking how different factors influence or don't influence children's behaviors.

However, the introduction does not make it clear how this study is the same or different than previous work. What is compassionate behavior and how is it different than instrumental helping? Assuming the authors are interested in cognition related to compassion: What makes responding in a Warnecken task different than their task? Is it just the presence of suffering? If this is the question of interest, then it is a bit odd that the tasks do not vary whether the puppets respond in distress. I am happy to see replications in this paper, but it seems a bit odd to introduce the paper as the first of its kind. The basic idea seems to be better tested in the Green et al paper where they varied whether the puppets display distress or not. This paper tests a different question: given that children do act compassionately what factors influence their behavior? Thus I would suggest that the authors make some big changes in the introduction to make this clear. In other words, the current experiments do not test what role the distress is playing in any of the children's behavior because it is not systematically varied.
Some more minor comments
[Note the page numbers are from the top of the document, not the ones printed on the page.]
1. It is not clear what the computational or representational difference is between ‘helping’ and ‘compassion’. Is a compassionate versus a prosocial behavior simply helping adopting someone else’s goal with the acknowledgment of their emotional state? Is there evidence that children do indeed acknowledge the emotion and wish to assuage it?
2. Page 2; Line 41: Unclear what skills children need?
3. Page 3, line 3, typo ‘by’ should be ‘be’
4. It’s very typical to say that this will help us design intervention programs. But how exactly?
5. Page 5 line 33 – is empathy-related responding the same as compassion?
6. Page 6 33 – ‘reassuring/friendly inner voice tones and primes seem to be experimental manipulations not facilitators in the abstract sense that I would expect to be in a ‘compassionate algoritm’. In other words what are these meant to operationalize?
7. Page 7 line 3 Olson Spelke did not find that, their study was a about third-party inferences/predictions, not how children engage with their own families. In that paragraph please be clear about what are first person behaviors and third-person predictions.
8. Page 7 line 54 – it is very weird to start this paragraph this way and then say that you will replicate part of another study that tested compassionate responding. That is, wasn’t the point of the Green et al experiment if not to test whether children act with compassion?
9. Page 9 line 19 – a bit more details should be given about each task or else collapse them. No one will know what the ‘Marble Task’ is.
10. Page 10/37; lines 10-45. It is not clear how ‘cost’ plays out in this experiment. Did the children have to give up some of their stickers to help? If its just about whether they got rewards, then the authors should rename the study. Ok I see this comes later. But it is very confusing how the stickers came into play.
11. Figure 2 – I’m not sure what this is meant to convey beyond Figure 1. And it is confusing that both bar colors are not called out, and that it changes from how information is displayed in Figure 1.
12. Totals in Consoling/Disengaging should be 31/X otherwise its meaningless.
13. Reference 47 could also Asaba, M., Li, X., Yow, W. Q., & Gweon, H. (2019). A friend, or a toy? Four-year-olds strategically demonstrate their competence to a puppet but only when others treat it as an agent. In CogSci (pp. 98-104).
14. Page 23 line 6 typos ‘adults and puppets’
15. Discussion: I am not sure why the explicit instructions is related to it being a competitive setting?
16. Citations for ‘known to be individualistic’. I’d also rewrite this to be more specific than ‘WEIRD’ do we really think that all other cultures would respond differently than all WEIRD cultures. In other words, I would just say something like, these were all [specific demographic].
17. Conclusion: In reading this I’m questioning whether it is important that children are compassionate. I’m still not convinced how understanding children’s compassion can help us build a more compassionate world. I would take out this statement or be more specific.

Thank you for addressing the concerns I raised in the original review.