Michael Overton
@moverton.bsky.social
Associate Professor of Public Administration at the University of Idaho
Associate Director of the Institute for Interdisciplinary Data Science
Data junkie in my professional life. Huge nerd in my personal life.
All opinions are my own.
Associate Director of the Institute for Interdisciplinary Data Science
Data junkie in my professional life. Huge nerd in my personal life.
All opinions are my own.
Yes, that was intentional.
June 13, 2025 at 9:13 PM
Yes, that was intentional.
It only suggested a small twerk.
June 13, 2025 at 9:12 PM
It only suggested a small twerk.
Are you sure there isn't a more timely topic? s/
March 17, 2025 at 7:56 PM
Are you sure there isn't a more timely topic? s/
It is pretty disingenuous to claim that you were always saying "rectangles are not GENERALLY squares" when you literally said "Entity classification is *not* a generative task." I really wish you had approached this in good faith.
March 16, 2025 at 12:26 AM
It is pretty disingenuous to claim that you were always saying "rectangles are not GENERALLY squares" when you literally said "Entity classification is *not* a generative task." I really wish you had approached this in good faith.
Are you discussing storytelling with data? If so, I have some slides that might be useful.
March 14, 2025 at 6:17 PM
Are you discussing storytelling with data? If so, I have some slides that might be useful.
The only thing I can think is that they were actually using a locally hosted LLM and was sending each row as a separate call.
March 14, 2025 at 6:04 PM
The only thing I can think is that they were actually using a locally hosted LLM and was sending each row as a separate call.
Like Dr. Resh, I am appreciative of good faith criticisms. But, I must disagree with your assessment. Generative AI especially reasoning models can be used in classification tasks. The LLM-as-a-judge lit is predicated on LLMs producing judgements often in the form of classifications.
March 14, 2025 at 6:02 PM
Like Dr. Resh, I am appreciative of good faith criticisms. But, I must disagree with your assessment. Generative AI especially reasoning models can be used in classification tasks. The LLM-as-a-judge lit is predicated on LLMs producing judgements often in the form of classifications.
Respectfully, the example would use multiple generative AI skills such as entity recognition and reasoning to infer threat levels. The example implies the use of a multi modal generative AI system, but their are corollaries with LLM natural language understanding and natural language inference.
March 14, 2025 at 4:59 AM
Respectfully, the example would use multiple generative AI skills such as entity recognition and reasoning to infer threat levels. The example implies the use of a multi modal generative AI system, but their are corollaries with LLM natural language understanding and natural language inference.