Hello, welcome to my talk page!

If you want to leave a message, please do it at the bottom, as a new section, for better formatting. You can do that by simply pressing the plus sign (+) or "new section" on the top of this page. And don't forget to sign your messages with four tildes, like this: ~~~~

Attention: I prefer to keep discussions unfragmented. If you leave a comment for me here, I will most likely respond to it on this same page—my talk page—as an effort to keep the entire conversation in one place. By the same token, if I leave a comment on your talk page, please respond to it there. Remember, we can use our watchlist and topic subscriptions to keep track of when responses are made. At the same time, feel free to send an alert to me on this page about a comment you have left elsewhere.

Thank you!

Computer Vision and Generative AI

Hello MrOllie,

Until yesterday, I also believed that image generation was not part of Generative AI. I started looking into the issue, and I found out that things might be changing (see for example this paper). What do you think? Thanks! LazyRabbitBlue (talk) 00:44, 13 March 2025 (UTC)[reply]

I think that an unpublished arxiv preprint isn't a usable source, and even if it were it wouldn't justify putting a sentence in the lead section of an article as you did. More generally, using generative AI to train a computer vision model results in a vision model that only works well on fake images. MrOllie (talk) 01:13, 13 March 2025 (UTC)[reply]
I see your point. I will think about it and maybe add a reference to Gen AI in a more discrete location. To be fair, Gen AI can be used to improve the quality of real image and thus help the following image processing steps. Examples include super resolution. LazyRabbitBlue (talk) 01:19, 13 March 2025 (UTC)[reply]
Artificially increasing the resolution of an image before you feed it into a vision system just makes the vision system run slower. MrOllie (talk) 01:23, 13 March 2025 (UTC)[reply]
True, but speed is not the only criterium applicable when designing a computer vision system. Sometimes you just have a few images, so increasing the quality of the initial information might be a good idea. LazyRabbitBlue (talk) 01:27, 13 March 2025 (UTC)[reply]
Those algorithms don't actually increase the quality, though. They look better to human eyes but not to a vision system. But this is all irrelevant to building an encyclopedia. MrOllie (talk) 01:33, 13 March 2025 (UTC)[reply]
OK. Anyway, people are starting to talk of Gen AI in CV. Somehow, this should be included in the page. I will do my best to do it. I know this is debatable but the change is happening, and I think readers should have some material on the topic. LazyRabbitBlue (talk) 01:37, 13 March 2025 (UTC)[reply]
Wikipedia is not a place to tell people about the latest happenings or bleeding edge research (see WP:NOT). This stuff only belongs on an encyclopedia when it is old, boring, settled science. Once that happens, if you have reliable sources you should follow up at the article's associated talk page rather than my user talk. MrOllie (talk) 01:38, 13 March 2025 (UTC)[reply]

March 2025

We appreciate your contributions, but in one of your recent edits to Naver, it appears that you have added original research, which is against Wikipedia's policies. Original research refers to material—such as facts, allegations, ideas, and personal experiences—for which no reliable, published sources exist; it also encompasses combining published sources in a way to imply something that none of them explicitly say. Please be prepared to cite a reliable source for all of your contributions. You can have a look at the tutorial on citing sources. Thank you. Yamamomo kakijiroo (talk) 14:48, 13 March 2025 (UTC)[reply]

Just stop edit warring and evading your block, you aren't fooling anyone. MrOllie (talk) 14:50, 13 March 2025 (UTC)[reply]
Information icon Hello, I'm Yamamomo kakijiroo. I noticed that you added or changed content in an article, Naver, but you didn't provide a reliable source. (Korea IT times) It's been removed and archived in the page history for now. If you think I made a mistake, you can leave me a message on my talk page. Thank you. Yamamomo kakijiroo (talk) 15:49, 13 March 2025 (UTC)[reply]
Nonsense warnings aren't going to suddenly make me think you aren't evading your block. MrOllie (talk) 15:51, 13 March 2025 (UTC)[reply]
Now blocked as "very likely", I see. Bishonen | tålk 20:25, 14 March 2025 (UTC).[reply]

Fortran book

Dear MrOllie,

I was notified that my revisions of Fortran has been reverted by you, and I would like to know the reasons. With these revisions I added the books by Metcalf, et al. in the Further Reading section of the article. Please refer to https://en.wikipedia.org/w/index.php?title=Fortran&oldid=1279677550 for the version after my revisions.

You commented "Restored revision 1278899946 by Xose.vazquez (talk): Rm books frequently spammed by SPAs" on your revert. What are Rm and SPA?

I am concerned that you regarded my editing activity as spamming. I wish to resolve any misunderstanding.

I saw that you have reverted revisions that added books by Metcalf, et al., several times in the past. If you have concerns about these particular books, I wish to know about them.

Thank you for your contributions to Wikipedia. I hope to work together with you and many others to develop a free encyclopedia.

Sincerely, Norio NorioTakemoto (talk) NorioTakemoto (talk) 21:46, 13 March 2025 (UTC)[reply]

'Rm' is removed and SPA is 'Single purpose account'. We see a lot of very new accounts trying to add those books to Wikipedia. Some of the additions are from accounts associated with the author, and others are simply accounts that have been asked to add the books on the author's behalf. My concern about these particular books is that Wikipedia should not be used as an avenue to promote them. MrOllie (talk) 22:36, 13 March 2025 (UTC)[reply]
Dear MrOllie,
thank you very much for your reply. I understand your concerns.
I hope you can see that my account is not a SPA from my edit history. I have never been asked by any of the authors to promote the book. I learned Fortran 90/95 from one of the two books to the extent that I can work as a computational physics researcher. I genuinely think that it's a useful book for many readers, and I also think that it's been found good by many.
There are already some books cited in the Fortran article. Is there an official rule about what books to cite and what not in the Further reading section?
Citing a book in an article will indeed automatically promote that book unless a negatively critical comment is attached to the citation. Those books currently in the article have been already promoted in this sense. Do you feel they should also be removed altogether?
How do you think about bringing up this topic in the Talk page of Fortran article?
Sincerely,
Norio NorioTakemoto (talk) 02:37, 14 March 2025 (UTC)[reply]

Reinforcement learning

Thank you for your alert. I think your complain is about "behaviour" instead of "behavior", and "modelled" instead of "modeled". I corrected it again. And I put my explanations back (sorry I had troubles with Wikipedia's visual editor, now it is ok). Fschwarzentruber (talk) 15:28, 14 March 2025 (UTC)[reply]

Allikas: http://en.wikipedia.org/wiki/User:MrOllie
No tags for this post.