The N64 GoldenEye watch theme.

But would it ruin it after a long wait?

Edit: Just started listening to a 30min version. Not ruined.

If you know or find any preprint servers people in your line of work use, feel free to add them to the issue, that would help me a lot.

Definitely not my line of work. I’m on the tools - https://www.explainxkcd.com/wiki/index.php/2501:_Average_Familiarity

The person that wrote the github issue seemed like they were somewhat aware of sources. Perhaps you could address the comment regarding best place to contact? A dedicated Lemmy community perhaps, spread the fediverse love?

I understand your hyperfocus for the app, because I assume most researchers or PhD candidates are also hyperfocused. I figured that news releases regarding the field of research is also relevant in some way, finger on the pulse sort of thing. Probably not as easy as grabbing the papers as you currently are.

https://github.com/jakobkreft/Aftergleam/issues/3

I can see that the crux of my problems drilling down on environmental topics may be made more difficult by some missing sources as illustrated in this issue. I’m still training away and getting some good hits so it’s working regardless.

Your app has a lot of potential. I’m going to use it as a positive/factual news source for a while.

Examples that are interesting to me by selecting 2 initial biology and ecology topics and then adding keywords (“soil” and “biochar”):

Coarse Woody Debris of Mixed Tree Species Overrides Soil Matrix Constraints on Microbial Richness and Abundance in an Ancient Broad-Leaved Forest https://www.biorxiv.org/content/10.64898/2026.10.09.757649

Found with Aftergleam, a privacy reader for arXiv scientific papers.

Synergistic effects of biochar and vegetation on microplastic and nanoplastic removal from stormwater runoff in bioretention systems https://doi.org/10.26434/chemrxiv.15010823/v1

Found with Aftergleam, a privacy reader for arXiv scientific papers.

Like I said, I’m not a scientist so I’m using keywords to hit terms that are relevant to me like “tree”, “species”, “biochar”, “plant”, “soil” etc. Then I deleted some after training and added “Australia” “forest” to train more relevant to my location to see what would happen. Then I might add my climate type later to drill further down.

So I can’t really search to find specific papers as I don’t technically know what I’m looking for. I’m extrapolating papers that might be relevant to my environmental line of work. Edit: Searched some individual keywords and I understand what you mean, it’s just time to select the papers that are relevant in that particular keyword. I’ll give that a try for a while and keep my keywords to broaden the feed.

If you were going to develop the app further, it could become quite a nice mixed feed reader for specific topics of interest, maybe with a sharing intent for social media too. I could imagine a News tab separate to For You and Explore that picked up scientific news releases based on keywords as well coming from the main journalistic sites like phys.org, theconversation, mongabay etc.

Thanks for your help.

With further keyword tweaking, I can get more relevant papers. As the keywords have a limit, if I train it with a certain list of keywords, then delete them and add new ones for further training, is this a way around that limit? Will the training stick for deleted keywords?

I don’t know enough but I feel as though this app will be very helpful for the people that require it. Well done.

I’m not a scientist, nor do I have a “field” but I do like what I like and it managed to find somewhat relevant papers in one go (it complained about it though) and that’s without further tweaking in the settings.

Very nice. Clean app too.

Don’t forget that someone has gotten up from bed and is wobbling over them trying to take a picture before they move, right at breakfast time.

But you’re probably right. Sometimes they like each other.