Artisobulus quoted the Pentateuch in Greek a few times, and so scholars have researched how closely these citations conform to the LXX. Usually, from what I've seen, his citations are considered to be pretty close to the LXX, but not necessarily exact. A new article by Sean Adams takes up the question and does a pretty good job of laying out the evidence.
The citations include the following:
- Exodus 13:9
- Exodus 3:20
- Exodus 9:3
- Exodus 20:11
- Deuteronomy 4:11
- Genesis 1
- Exod 13:9: the LXX has ἐξήγαγέν σε κύριος while Aristobulus has ἐξήγαγέν ὁ θεός σε.
- Exod 3:20: the LXX has καἰ ἐκτείνας τὴν χεῖρα πατάξω τοὺς Αἰγυπτίους while Aristobulus has ἀποστελῶ τὴν χεῖρά μου καὶ πατάξω τοὺς Αἰγυπτίους.
- Exod 20:11: the LXX has the subject κύριος while Aristobulus does not express a subject, along with a few other differences (p. 7 of the article).
The examination of the evidence here is helpful, but I doubt that the solutions offered are really to the point. Aristobulus possibly knew Hebrew, but I think scholars are correct to say that there is very little evidence for this position. His quotations do not provide such evidence if a more likely explanation can be found for their closeness to the MT as opposed to the LXX. And the other suggestion, that Aristobulus had access to a non-Septuagintal Greek text seems to be an unacknowledged revival of the theory argued vigorously by Paul Kahle such that there was no single original Greek translation but a variety of translations that coalesced later on. For the past several decades, LXX specialists have regarded Kahle's view as not the best way to understand the origins of the LXX. (See, e.g., Fernández Marcos, 53-57.)
Probably the best explanation for these citations by Aristobulus is that he has access to a Greek text of Exodus that has been revised toward the proto-MT text form. The reason this seems to be the best explanation is that such Greek texts are so common in the manuscript tradition. (See, e.g., T.M. Law, ch. 7, which Adams has recently reviewed.) This is the explanation offered by Barthélemy (332 n. 24). The evidence seems to be so minor and sporadic in this case that it is possible to argue that such a revision would have been unintentional, accomplished by someone so familiar with the Hebrew text that he (or she) unconsciously revised the Greek text to accord with it. This is the explanation I proposed in my book (pp. 158-59), echoing an idea from J. W. Wevers.