Innovative topic and research design with promising clinical implications. I commend the authors for addressing this understudied part of the field in bipolar disorder/therapy. I have some concerns regarding the methodology and study design. I am aware that you cannot change recruitment/overall design but I would encourage the author to either re-think some of the analysis and/or be more insightful when describing expected outcomes of the study and when interpreting their results. Please address the comments here below in your paper, either in the introduction or limitations. and consider ways to attenuate the effects of theses biases in your analyses.

1. including BD (of any type, including NOS) and MDD in the same design add a high lever of heterogeneity to this type of study. the nature of depressive episodes is simply different, both in terms of biological implications, thinking style, and effect on global functioning. BD is also associated with a number of comorbidities. I think the authors should address this in their background and when developing hypotheses. please include statistics regarding the number of BD-spectrum vs MDD in both therapeutic groups. please let the reader know how many BD type 1 participants you had. This group may show additional executive function impairment/impulsivity during episodes.
   1b. please provide a short description of cognitive profile in MDD and BD.
   1c. did you consider running your analyses including either only BD or only MDD? this could reduce statistical noise/variability.
2. would like to know more about number of episodes, onset of the disease, suicidal risk, family history of disease in both groups. please include F, p values, odd ratios/effect size to provide this overview
3. global cognition. it is well known the that attentional fluctuations, working memory and fluctuating learning curve on RAVLT are primary features of mood disorders/anxiety. however there is literature regarding the stronger effect of mood on memory in general in BD. In BD these deficits often go back to baseline in between episodes. This depends on the severity of the disease of course. this means that some of your participants may have recovered from an episode "naturally". so expecting CR to simply improve their cognitive performance. since you included all BDs and MDDs together I wonder if it is a valid enough approach to expect cognitive improvement in both groups simply by providing CR. one way to approach this could be either to exclude individuals with multiple episodes or only include individuals that already showed good management of their BD symptoms. your groups will be otherwise too heterogenous at the biological and cognitive level. and your results will be challenging to interpret in a larger sense.
   in Table 1 please include all demographic, episode-related and biology related findings and please provide stats to allow us to compare your groups.
4. creating a composite score of cognition is too broad. i would focus on one or two cognitive domains of interest eg attention span, immediate memory that have been found to be changing during mood episodes
5. BD is often comorbid to ADHD, OCD, Substance use etc. please provide details on these comorbidities. they do play a role on cognitive performance and response to treatment overall. also please let the reader know re trauma symptoms and other personality traits that could affect response to CR.
6. did you balance the two therapeutic groups in terms of gender (based on statistical prevalence of MDD and BD in females/males/non-binary)
7. please provide some background in terms of global functioning, how did you recruit participants, were they employed, if they decided to exit therapy for this study how was that managed?
8. please describe the background of therapists providing IPSRT and CR in groups, psychologist, researcher psychologist, psychiatrist etc.
9. could you please include data on suicidal risk, did you include this in your outcome measures?
10. did you consider including a control group?
11. please provide stats in all your tables to compare groups. I need more data to make sense of your results.especially b values, odd ratios, R2 for your regressions. how did you handle missing data, multicollinearity, which covariates did you include based on differences between groups?

I am satisfied with the authors' revisions.