Uploaded November 2024 | Updated November 2026, 4 days ago
Abstract: We present Atlantes, the first global-scale, real-time system for analyzing vessel behavior through GPS data. Our system integrates a hierarchy of novel transformer-based models that precisely classify behaviors from noisy and irregular spatiotemporal GPS sequences. We outline our approach to generating a large, high-quality training dataset (~15 million expertly annotated GPS messages), the model architecture and machine learning strategy, and a staging environment which enables fast model iteration informed by real-world feedback. Currently deployed for maritime intelligence applications within the Skylight platform, the model processes over 2.6 million trajectories (~3 billion GPS messages) daily. This will be a technical presentation with relevance to both the research and engineering communities.
Bio: Henry Herzog is a Machine Learning Engineer within the Applied Sciences Organization at Ai2. He joined Ai2 following a summer internship with the Skylight team, where he was recognized as an Outstanding Intern in 2023 for his contributions to developing novel deep learning methods for behavior analysis using GPS sequences. He holds a BA in Mathematics from the University of Chicago (2023).
Abstract: We present Atlantes, the first global-scale, real-time system for analyzing vessel behavior through GPS data. Our system integrates a hierarchy of novel transformer-based models that precisely classify behaviors from noisy and irregular spatiotemporal GPS sequences. We outline our approach to generating a large, high-quality training dataset (~15 million expertly annotated GPS messages), the model architecture and machine learning strategy, and a staging environment which enables fast model iteration informed by real-world feedback. Currently deployed for maritime intelligence applications within the Skylight platform, the model processes over 2.6 million trajectories (~3 billion GPS messages) daily. This will be a technical presentation with relevance to both the research and engineering communities.
Bio: Henry Herzog is a Machine Learning Engineer within the Applied Sciences Organization at Ai2. He joined Ai2 following a summer internship with the Skylight team, where he was recognized as an Outstanding Intern in 2023 for his contributions to developing novel deep learning methods for behavior analysis using GPS sequences. He holds a BA in Mathematics from the University of Chicago (2023).






![Explaining Answers with Entailment Trees
Explaining Answers with Entailment Trees
Bhavana Dalvi
https://allenai.org/team/bhavanad
Our goal, in the context of open-domain textual question-answering (QA), is to explain answers by showing the line of reasoning from what is known to the answer, rather than simply showing a fragment of textual evidence (a rationale). If this could be done, new opportunities for understanding and debugging the systems reasoning become possible. Our approach is to generate explanations in the form of entailment trees, namely a tree of multipremise entailment steps from facts that are known, through intermediate conclusions, to the hypothesis of interest (namely the question + answer). To train a model with this skill, we created ENTAILMENTBANK, the first dataset to contain multistep entailment trees. Given a hypothesis (question + answer), we define three increasingly difficult explanation tasks: generate a valid entailment tree given (a) all relevant sentences (b) all relevant and some irrelevant sentences, or (c) a corpus. We show that a strong language model can partially solve these tasks, in particular when the relevant sentences are included in the input (e.g., 35% of trees for (a) are perfect), and with indications of generalization to other domains. This work is significant as it provides a new type of dataset (multistep entailments) and baselines, offering a new avenue for the community to generate richer, more systematic explanations. [ Paper published in EMNLP 2021, https://arxiv.org/abs/2104.08661 ] Explaining Answers with Entailment Trees](https://i.ytimg.com/vi/QPSZQYA1RmA/mqdefault.jpg)



