To check whether or not this drawback holds for immediately’s giant multimodal fashions, the group carried out a managed analysis. They skilled the chosen fashions on 5 goal duties, together with fine-grained chicken classification, counting, medical visible query answering, OCR studying, and time studying. They then measured how a lot efficiency dropped throughout eight commonplace benchmarks that weren’t a part of the fine-tuning set.
These experiments led to 2 key discoveries, in line with the paper. Tuning solely the self-attention projection layers (SA Proj), the a part of the mannequin that helps it resolve which enter components to give attention to, allowed the fashions to study new duties with little or no measurable forgetting. Additionally, what initially appeared as forgotten data typically resurfaced when the mannequin was later skilled on one other specialised job.
“We thus hypothesize that maybe what seems like forgetting or interference after fine-tuning on a slender goal job is definitely bias within the output distribution as a result of job distribution shift,” the researchers added. “By means of in-depth evaluation when tuning the counting job, we affirm this speculation: tuning the MLP will increase goal accuracy but additionally will increase the chance of outputting numeric tokens and a extremely correlated drop in held-out job accuracy, whereas tuning the self-attention achieves the goal studying with out a lot bias towards numeric tokens and with out shedding held-out accuracy.”
