Blake Lemoine

The latest edition of the Against the Grain Journal is hit inboxes today is all about AI in libraries! We're excited this issue was guest edited by Vessela Ensberg, Associate Director, Data Architecture at UC Davis Library, and Peter Brantley, Director of Online Strategy at UC Davis Library.     Inside you'll find • An interview with powerhouse Gary Price, discussing trust and transparency, citations and provenance, and literacy for using AI, among other things.  • Lucy Lu Wang, from the Information School and University of Washington, discussing generative AI for information access.  • Lorcan Dempsey, also from the iSchool at UW, tells us about Libraries, AI and Training Collections.  • Brandon Butler, recently named Executive Director at ReCreate Coalition, talks about some of the legal issues surrounding the training of generative technologies.  • Francesco Ramigni of the ACMI, Australia’s Museum of Screen Culture, tells us how text to- image/video generative AI may help libraries and museums collections accessibility, and includes some fascinating videos embedded as part of the article. • Shawn Averkamp from AVP celebrates embracing serendipitous innovation in libraries. From our regular columnists, we have another amazing installment of the Readers Roundup reviews, edited by Corey Seeman, Director of Kresge Library Services, Ross School of Business, University of Michigan. In addition to being famous for his squirrel photography side gig, Corey keeps us all informed with reviews of reference materials ranging from “I need this book on my nightstand” to “I’ll use my money elsewhere.” Corey’s editors note in this installment is especially poignant given the theme of this issue — be sure to give it a read! There’s lots more to unpack in this issue, but I’ll end it here and let you get to reading! Not signed up? Don't miss another issue! Subscribe online at: https://lnkd.in/eR3amum9

43

A Survey of RAG and RAU: Advancing Natural Language Processing with Retrieval-Augmented Language Models Quick read: https://lnkd.in/gnGqk44W Paper: https://lnkd.in/g8-VftbX #llms #largelanguagemodels #ai

20

So, there are six steps in the utils.py file... here's a summary of what the prompts do. PROMPT 1: Request an initial translation of the entire source text without explanations. PROMPT 2: Ask for constructive criticism and suggestions to improve the initial translation. PROMPT 3: Request an improved translation based on expert suggestions and criticisms. PROMPT 4: Request a translation of a specific chunk of the source text using the rest for context. PROMPT 5: Ask for constructive criticism and suggestions to improve a specific chunk's translation. PROMPT 6: Request an improved translation of a specific chunk based on expert suggestions. From Xuyun Zeng

3

What's the difference between model explainability, interpretability, and observability? Jason Zhong's debut TDS article outlines the nuanced distinctions you should be aware of.

16

1 Comment

Researchers at CDS have unveiled that transformer language models can solve complex problems without clear intermediate reasoning, challenging existing assumptions about their operation. Led by CDS PhD students Jacob Pfau and William Merrill, and CDS Associate Professor of Linguistics and Data Science Sam Bowman, this study highlights the use of nonsensical filler tokens that achieve perfect accuracy in specific tasks, raising concerns about  interpretability and accountability in LLMs. Explore the full findings: https://lnkd.in/eERzzKsU

13

Webinar alert - April 30, 1PM EDT. George Anadiotis from LinkedDataOrchestration and George Demarest of #Aerospike will guide you through selecting the right graph database and essential selection criteria tailored to your organizational needs. https://lnkd.in/eDV-QQuB #GraphDatabase #DataManagement

10

🚀 Ready to optimize your vector database operations? Dive into our latest blog post: "Tracing Weaviate API Calls with Langtrace!" Discover how to seamlessly integrate Langtrace to monitor and trace your Weaviate interactions, unlocking valuable insights and boosting performance. Elevate your data management game now! #Langtrace #Weaviate #DataOptimization 📊🔍

6

🧮 AI that can do math, now available on Flint! Large language models (LLMs) typically struggle when it comes to math. To overcome this, the AI on Flint now uses a powerful calculator under the hood to provide accurate results. 🤓🧠 Some other recent updates to Flint include: 1️⃣ A built-in equation editor to let students show their work on math problems. 2️⃣ Evidenced-based feedback to show exactly where a student struggled and where they excelled. 3️⃣ A grade and language level selector to set the difficulty of the AI. #ai #education #edtech #aieducation #k12

37

2 Comments

Just released a blog post about retrieval augmented generation (RAG), Ollama, and air-gapped AI. See how Canisius University is creating completely offline, personalized, AI pipelines. Heck - it even talks about Docker too! #ollama #airgappedAI #RAG

10

2 Comments

We're thrilled to share our latest study, "Prompt-Time Ontology-Driven Symbolic Knowledge Capture with Large Language Models," on arXiv. This research explores innovative methods to enhance large language models (LLMs) with the ability to learn personal information—directly from user prompts—using ontology and knowledge-graph approaches. By fine-tuning LLMs with a predefined ontology, we improve their capability to incorporate user-specific information efficiently, leading to more personalized interactions. Our findings indicate that this approach significantly advances the personalization of AI technologies, bringing us closer to developing systems that adapt to each user's unique needs. Find the full preprint at https://lnkd.in/d2qJrAxf or https://lnkd.in/dnUvBH9g

21

Join us on May 29, 2024, for a webinar covering deep learning frameworks on Argonne National Laboratory's Aurora exascale supercomputer. In this developer session, ALCF's Corey Adams will discuss deep learning frameworks optimized for Aurora, including TensorFlow, PyTorch, and the experimental release of JAX. Other topics include how to achieve optimal single device performance, as well as scaling up and scaling out models. The session will include examples and demonstrations that can run on Sunspot or Aurora. https://lnkd.in/gY2WMZ3x

7

This week's session will provide practical examples of using graph queries on system #digitalthreads to solve real-world engineering problems using the all-new Syndeia 3.6. Users will learn about key concepts in digital thread graphs, explore the new query library, formulate queries using the Syndeia query builder and Gremlin query language, and visualize results. Register here if not done already https://lnkd.in/diBz36wB Please email us at info@intercax.com for more information. #digitaltransformation #digitalthread #API #systemsengineering #MBSE #MBE #digitalengineering #training #syndeia36

2

In this blog post, we'll explore the various applications where RAG is particularly well-suited, including tasks requiring factual correctness and more. #DataScience #AI #ArtificialIntelligence https://hubs.li/Q02ttqHM0

1

In this blog post, we'll explore the various applications where RAG is particularly well-suited, including tasks requiring factual correctness and more. #DataScience #AI #ArtificialIntelligence https://hubs.li/Q02tr-PZ0

2

Become an expert on aiXplain’s no-code AI platform and SDK with our #aiXpertTrainingCourse 🤓 In this tutorial, learn to transform your linguistic corpora into structured datasets with aiXplain, enhancing your AI's understanding and performance. Don't forget to subscribe to our YouTube channel for more. 👉 https://lnkd.in/gHCmvZwn #aiXplain #DataScience #Corpus #AI #MachineLearning #LearnAI #AIplatform #AItools

7

Thanks to the North American Chapter of the Association for Computational Linguistics (NAACL), SIMBig 2023 awarded cash amounts for the outstanding papers presented at the conference. Congratulations to our four winners! #AI #datascience #nlp #artificialintelligence #BigData #healthinformatics #MachineLearning #ML #naacl

6