Aeneas Journal Club скачать в хорошем качестве

Aeneas Journal Club

Join the GDG AI for Science community to explore Aeneas, the artificial intelligence model from Google DeepMind designed to contextualise ancient inscriptions. Historically, identifying "parallels" in ancient texts has been a complex and time-consuming task for historians, especially when dealing with fragmentary or defaced inscriptions. Aeneas significantly accelerates this work by reasoning across thousands of Latin inscriptions, retrieving textual and contextual parallels in seconds. This allows historians to interpret and build upon the model's findings, providing richer insights into the diversity of everyday life in the Roman world. In this collaborative session, we will unpack the capabilities of Aeneas to understand: What Aeneas is: The first AI model designed to contextualise ancient inscriptions. How it works: Reasoning across vast collections of Latin inscriptions to find textual and contextual parallels. What Aeneas is capable of: Searching for parallels, processing multimodal input (text and images) for geographical provenance, and restoring gaps in texts. Aeneas has shown state-of-the-art performance in restoring damaged texts and predicting when and where they were written. How to use it: We can demo the freely available interactive version for researchers, students, educators, and museum professionals, as well as its open-sourced code and dataset for further research on your data. Its impact and limitations: How Aeneas aids in interpreting isolated fragments, draws richer conclusions, and offers a new quantitative way to engage with historical debates. Read the DeepMind blog post: https://deepmind.google/discover/blog... Read the Nature article: https://www.nature.com/articles/s4158... Explore the online demos in Latin (Aeneas) and Greek (Ithaca): https://predictingthepast.com/ Dive deeper into configuration options with the interactive Google Colab Notebook: https://colab.research.google.com/git... Leverage the "education focused" versions for teaching: https://colab.research.google.com/dri... Ask questions in out curated study guide on NotebookLM: https://notebooklm.google.com/noteboo... Join the GDG AI for Science community for talks, events, workshops, collaborations and more: https://gdg.community.dev/gdg-ai-for-...