2023-04-24-alazraki23a.md

title	abstract	layout	series	publisher	issn	id	month	tex_title	firstpage	lastpage	page	order	cycles	bibtex_author	author	date	address	container-title	volume	genre	issued	pdf	extras
How (not) to ensemble LVLMs for VQA	This paper studies ensembling in the era of Large Vision-Language Models (LVLMs). Ensembling is a classical method to combine different models to get increased performance. In the recent work on Encyclopedic-VQA the authors examine a wide variety of models to solve their task: from vanilla LVLMs, to mod- els including the caption as extra context, to models augmented with Lens-based retrieval of Wikipedia pages. Intuitively these models are highly complementary, which should make them ideal for ensembling. Indeed, an oracle experiment (Fig. 1) shows potential gains from 48.8% accuracy (the best single model) all the way up to 67% (best possible ensemble). So it is a trivial exercise to create an ensemble with substantial real gains. Or is it?	inproceedings	Proceedings of Machine Learning Research	PMLR	2640-3498	alazraki23a	0	How (not) to ensemble LVLMs for VQA	1	20	1-20	1	false	Alazraki, Lisa and Castrejon, Lluis and Dehghani, Mostafa and Huot, Fantine and Uijlings, Jasper and Mensink, Thomas	given family Lisa Alazraki given family Lluis Castrejon given family Mostafa Dehghani given family Fantine Huot given family Jasper Uijlings given family Thomas Mensink	2023-04-24		Proceedings on "I Can't Believe It's Not Better: Failure Modes in the Age of Foundation Models" at NeurIPS 2023 Workshops	239	inproceedings	date-parts 2023 4 24	https://proceedings.mlr.press/v239/alazraki23a/alazraki23a.pdf