No abstract

No abstract

Contrastively trained image-text models such as CLIP, ALIGN, and BASIC have demonstrated unprecedented robustness to multiple challenging natural distribution shifts. Since these image-text models differ from previous training approaches in several ways, an important question is what causes the large robustness gains. We answer this question via a systematic experimental investigation. Concretely, we study five different possible causes for the robustness gains: (i) the training set size, (ii) the training distribution, (iii) language supervision at training time, (iv) language supervision at test time, and (v) the contrastive loss function. Our experiments show that the more diverse training distribution is the main cause for the robustness gains, with the other factors contributing little to no robustness. Beyond our experimental results, we also introduce ImageNet-Captions, a version of ImageNet with original text annotations from Flickr, to enable further controlled experiments of language-image training.

Feature selection is of paramount importance for text-mining classifiers with high-dimensional features. The Turku Event Extraction System (TEES) is the best performing tool in the GENIA BioNLP 2009/2011 shared tasks, which relies heavily on high-dimensional features. This paper describes research which, based on an implementation of an accumulated effect evaluation (AEE) algorithm applying the greedy search strategy, analyses the contribution of every single feature class in TEES with a view to identify important features and modify the feature set accordingly. With an updated feature set, a new system is acquired with enhanced performance which achieves an increased F-score of 53.27% up from 51.21% for Task 1 under strict evaluation criteria and 57.24% according to the approximate span and recursive criterion.

This paper presents the DialogBank, a new language resource consisting of dialogues with gold standard annotations according to the ISO 24617-2 standard. Some of these dialogues have been taken from existing corpora and have been reannotated, offering the possibility to compare annotations according to different schemes; others have been newly annotated directly according to the standard. The ISO standard annotations in the DialogBank make use of three alternative representation formats, which are shown to be interoperable. The (re-)annotation brought certain deficiencies and limitations of the ISO standard to light, which call for considering possible revisions and extensions, and for exploring the possible integration of dialogue act annotations with other semantic annotations.