Search for: All records

When reading narratives, human readers rely on their Theory of Mind (ToM) to infer not only what the characters know from their utterances, but also whether characters are likely to share common ground. As in human conversation, such decisions are not infallible but probabilistic, based on the evidence available in the narrative. By responding on a scale (rather than Yes/No), humans can indicate commitment to their inferences about what characters know (ToM). We use two prompting approaches to explore (i) how well LLM judgments align with human judgments, and (ii) how well LLMs infer the author’s intent from utterances intended to project knowledge in narratives. 

ABSTRACT Here, we report on the raw and coassembled metatranscriptomes of 39 Lake Erie surface (1.0 m) water samples collected over a 2-day diel period encompassing episodic weather and bloom events. Preliminary taxonomic annotations and read mappings revealed thatMicrocystisspp. accounted for up to ~47% of the transcriptionally active community. 

Abstract Flowers can be transmission platforms for parasites that impact bee health, yet bees share floral resources with other pollinator taxa, such as flies, that may be hosts or non-host vectors (i.e., mechanical vectors) of parasites. Here, we assessed whether the fecal-orally transmitted gut parasite of bees,Crithidia bombi, can infectEristalis tenaxflower flies. We also investigated the potential for two confirmed solitary bee hosts ofC. bombi,Osmia lignariaandMegachile rotundata, as well as two flower fly species,Eristalis arbustorumandE. tenax,to transmit the parasite at flowers. We found thatC. bombidid not replicate (i.e., cause an active infection) inE. tenaxflies. However, 93% of inoculated flies defecated liveC. bombiin their first fecal event, and all contaminated fecal events containedC. bombiat concentrations sufficient to infect bumble bees. Flies and bees defecated inside the corolla (flower) more frequently than other plant locations, and flies defecated at volumes comparable to or greater than bees. Our results demonstrate thatEristalisflower flies are not hosts ofC. bombi, but they may be mechanical vectors of this parasite at flowers. Thus, flower flies may amplify or diluteC. bombiin bee communities, though current theoretical work suggests that unless present in large populations, the effects of mechanical vectors will be smaller than hosts. 

Abstract Classifying images using supervised machine learning (ML) relies on labeled training data—classes or text descriptions, for example, associated with each image. Data‐driven models are only as good as the data used for training, and this points to the importance of high‐quality labeled data for developing a ML model that has predictive skill. Labeling data is typically a time‐consuming, manual process. Here, we investigate the process of labeling data, with a specific focus on coastal aerial imagery captured in the wake of hurricanes that affected the Atlantic and Gulf Coasts of the United States. The imagery data set is a rich observational record of storm impacts and coastal change, but the imagery requires labeling to render that information accessible. We created an online interface that served labelers a stream of images and a fixed set of questions. A total of 1,600 images were labeled by at least two or as many as seven coastal scientists. We used the resulting data set to investigate interrater agreement: the extent to which labelers labeled each image similarly. Interrater agreement scores, assessed with percent agreement and Krippendorff's alpha, are higher when the questions posed to labelers are relatively simple, when the labelers are provided with a user manual, and when images are smaller. Experiments in interrater agreement point toward the benefit of multiple labelers for understanding the uncertainty in labeling data for machine learning research.