Our notes from reading. Instruction Tuning for Large Language Models: A Survey.
IPTC Media Topics
The IPTC (International Press Telecommunications Council) calls itself the global standards body for news media. In this article we focus on one of its standards: Media Topics. This standard contains a subject taxonomy for media, which is a topic hierarchy that can be used to classify news articles. *Example of a subset of Media Topics…
Apify, Geneea, and Google reviews
Are you curious about the aspects of iconic European landmarks that are disliked by some tourists? Dive into Apify’s blog post for a candid look at what visitors hate. Plus, don’t miss the chance to harness insights from Google reviews using our powerful text analyzer on the Apify platform. Are you ready to uncover some…
Geneea’s AI Spotlight #4
The fourth edition of our newsletter on Large Language Models is here.
Today, we look at
• Llama 2, the new model from Meta,
• practical aspects of LLM use – new tools but also some challenges,
• use of AI in media,
• and more.
Geneea’s AI Spotlight #3
The third edition of our newsletter on Large Language Models is here.
Today, we look at
• an introduction to LLM models by Andrej Karpathy;
• two posts on practical aspects of using LLMs; and
• the regulation of AI by the EU.
Geneea’s AI Spotlight #2
The second edition of our newsletter on Large Language Models is here. Below are summaries of papers and posts that captured our attention most during the last two weeks.
Today, we look at:
• various practical challenges and how to address them,
• two new models: Google Palm 2 and Falcon,
• LIMA and its approach to fine-tuning, and
• we again mention several non-technical LLM topics.
Geneea’s AI text analyzer in Apify
Looking to uncover valuable insights from Google Reviews about your business or your competitors? Look no further! Geneea has integrated its own actor into Apify’s web scraping and automation platform. Our actor is designed to analyze Google reviews, providing you with tags, attributes, and sentiment for each review. It is easy to visualize the resulting…
Geneea’s AI Spotlight newsletter
Geneea is a text analytics company, and AI has always been an integral part of what we do. Generative AI, transformers, and especially language models have been around for some time, but their popularization by OpenAI’s ChatGPT has unlocked immense creativity and sparked debates all over the world. We’re no exception. Not a day goes…
New version of Frida
Winter hibernation is coming to an end and we’re ready to tell you what’s cooking in our NLP kitchen: a new, slick version of Frida – a web application that provides comprehensive visualization of document and article analysis. We’ve incorporated valuable feedback from our clients, gathered over the years. We’ve improved UX, increased customizability, and…
News Impact Summit 2022
Geneea and the Czech News Agency discussed their experiences of using automatic text generation in election coverage at this year’s News Impact Summit: The Future of Editorial. The event took place in Prague on October 6, 2022 and was organized by the European Journalism Centre and the Google News Initiative to celebrate innovation in journalism…
KPMG Data Festival 2022
The 5th KPMG Data Festival, which took place on October 14, 2022 in Prague, was a great success. We sincerely thank the organizers for putting together such an ambitious program and enabling so many data enthusiasts to meet and share their experiences. We certainly had fun! In the numerous talks, experts introduced various ways in…
Covering elections 2022
Recently, and for the third time in a row, Geneea collaborated with Czech News Agency (CTK) in covering elections in the Czech Republic. On this occasion it was the 2022 local and senate elections. Our NLG automatically generated over 500 reports announcing preliminary and then final election results as they came in. This was done…
Geneea’s NLP in textbook preparation
Thinking of living in Prague for a while? The latest textbook of Czech for foreigners, Čeština EXPRES Start, has just been published. Geneea is proud to have taken part in its creation. We previously provided our NLP services to the popular set of language textbooks, Czech Step by Step, and to a simplified version of…
Václav Moravec on media, AI, and Geneea
Václav Moravec is a popular Czech columnist and TV host. As a guest of the podcast Welcome the Future, he discusses problems with contemporary journalism, how it must change, and its future. To illustrate what AI has to offer to media publishers, he mentions the long-term cooperation between Geneea and the Czech News Agency on…
Geneea on TV
Last month we welcomed television TA3, a privately-owned Slovak news channel, to our Prague offices! On June 27, 2022, they came to film us at work and collect information about our product for a news report that was later featured in their program, The World of Technologies. We spent a pleasant morning with them. It was…