True. It would be interesting to run these same tests on the 7B model to see if trend information changes or not. 7B had a march cutoff so if the aug-dec dip mi

True. It would be interesting to run these same tests on the 7B model to see if trend information changes or not. 7B had a march cutoff so if the aug-dec dip migrated to oct-march (or just disappeared) it would be strong evidence for training/data bias. If nothing else, comparing 7B to 70B would likely be interesting. edit I realized too late I had the years off. It is pure coincidence of month, not a real data bias. Sorry! I still think it would be interesting to see a 7B comparison but that is just to see how well a small model could spot big trends compared to a bigger one.