The lectures moved fast, so at times students were scrambling to keep up/write stuff down. The average for the first midterm was a 35/40, and was in the lower B range for the second midterm and final I think. Her exams were generally reasonable, the hardest part was the True/False, which required a deeper understanding of concepts and applications of theorems that one would not know unless they had spent a significant amount of time working through many problems (and even then, one might not completely get the problem). There was an issue with how she graded the second midterm's true/false section, giving 2 points for a correct answer, 1 point for no answer, and -1 points for an incorrect answer. According to her, the reason behind this was to discourage people from simply choosing an answer without truly knowing the reason behind it, but it actually punished the people who may have put effort into studying but just arrived at the wrong answer (tricky questions) and benefited those who may not have put in the effort to study, did not know the answer, and therefore just left it blank. There is no way of distinguishing those who simply circled an answer for the sake of guessing and possibly earning points from those who knew their shit and just arrived at the wrong answer for any other reason. Still, the final was reasonable if one truly knew the material, and she was very generous with the final grading (generally rounding people who were in between grades up; the grade distribution was better than that of the average math class at UCLA). Out of the homework that Bhaskar assigned, only a few (generally trickier problems that required a lot of thought) were mandatory to turn in weekly. No quizzes. Would recommend!