Denis Mongin
@denismongin.bsky.social
Started with lasers and nonlinear optics, now with biomedical informatics, data and statistics in health sciences.
Research in #Metaresearch, #rheumatology and #publichealth at HUG and University of Geneva.
#Rstats ninja, #Openscience
Research in #Metaresearch, #rheumatology and #publichealth at HUG and University of Geneva.
#Rstats ninja, #Openscience
LLM are incredible tools to deal with data extraction and multi label classification of unstructured text. Way above any other tools, without needing specific domain training.
And there are open weight LLM you can run locally to do so.
And there are open weight LLM you can run locally to do so.
December 4, 2025 at 10:21 PM
LLM are incredible tools to deal with data extraction and multi label classification of unstructured text. Way above any other tools, without needing specific domain training.
And there are open weight LLM you can run locally to do so.
And there are open weight LLM you can run locally to do so.
Magnifique, bravo
December 3, 2025 at 6:36 AM
Magnifique, bravo
Like the majority of us.
Who don't write much either here
Who don't write much either here
December 2, 2025 at 6:23 AM
Like the majority of us.
Who don't write much either here
Who don't write much either here
Alors non, et ça vaut le coup de prendre une heure pour écouter un spécialiste du sujet histoire d'avoir une comparaison correcte:
www.youtube.com/watch?v=6HjH...
www.youtube.com/watch?v=6HjH...
Les Vraies Raisons derrière la Montée du Fascisme - Johann Chapoutot
YouTube video by Metabolism of Cities
www.youtube.com
December 1, 2025 at 10:07 AM
Alors non, et ça vaut le coup de prendre une heure pour écouter un spécialiste du sujet histoire d'avoir une comparaison correcte:
www.youtube.com/watch?v=6HjH...
www.youtube.com/watch?v=6HjH...
Scientific report is publishing garbage since some years already. And a lot.
I realized by being a reviewer once for a super bad study, asking for rejection, and seing the article published anyway.
I realized by being a reviewer once for a super bad study, asking for rejection, and seing the article published anyway.
November 30, 2025 at 8:45 PM
Scientific report is publishing garbage since some years already. And a lot.
I realized by being a reviewer once for a super bad study, asking for rejection, and seing the article published anyway.
I realized by being a reviewer once for a super bad study, asking for rejection, and seing the article published anyway.
Just for the sake of it, a reminder that iceberg do not look like that, because they would not float like that.
Go try it:
joshdata.me/iceberger.html
Go try it:
joshdata.me/iceberger.html
November 30, 2025 at 6:51 PM
Just for the sake of it, a reminder that iceberg do not look like that, because they would not float like that.
Go try it:
joshdata.me/iceberger.html
Go try it:
joshdata.me/iceberger.html
If you try it your brain makes you forget what was done to produce it
November 28, 2025 at 9:25 PM
If you try it your brain makes you forget what was done to produce it
It is not that unexpected: I have seen many cases where reviews are not properly done: reviewers barely read the article, find the point in methods they care about, and get one of their article cited.
Reviewing takes time, that we don't have or that we don't want to spent for 35% profit margins
Reviewing takes time, that we don't have or that we don't want to spent for 35% profit margins
November 28, 2025 at 9:13 PM
It is not that unexpected: I have seen many cases where reviews are not properly done: reviewers barely read the article, find the point in methods they care about, and get one of their article cited.
Reviewing takes time, that we don't have or that we don't want to spent for 35% profit margins
Reviewing takes time, that we don't have or that we don't want to spent for 35% profit margins
scientific report has been such a disaster these last years....
I have seen so many bad studies there.
What a pity
I have seen so many bad studies there.
What a pity
November 27, 2025 at 1:05 PM
scientific report has been such a disaster these last years....
I have seen so many bad studies there.
What a pity
I have seen so many bad studies there.
What a pity
ou c'est un/des emplous fictifs: ils ne fait pas c'est quelqu'un d'autre qui fait mais il touche la paye
November 27, 2025 at 10:12 AM
ou c'est un/des emplous fictifs: ils ne fait pas c'est quelqu'un d'autre qui fait mais il touche la paye
I think DOGE hasn't been about cutting federal budget, but about getting access to key data and lever within the state while weakening agencies.
Now that it's done, it has no reason to exist
Now that it's done, it has no reason to exist
November 23, 2025 at 2:26 PM
I think DOGE hasn't been about cutting federal budget, but about getting access to key data and lever within the state while weakening agencies.
Now that it's done, it has no reason to exist
Now that it's done, it has no reason to exist
Oui, la vie est belle pour le moment à Genève.
Si un jour tu passes, dis moi !
Si un jour tu passes, dis moi !
November 19, 2025 at 2:57 PM
Oui, la vie est belle pour le moment à Genève.
Si un jour tu passes, dis moi !
Si un jour tu passes, dis moi !
"Among 1792 manuscripts in which the DAS indicated that authors are willing to share their data, 1669 (93%) authors either did not respond or declined to share their data with us"
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
Many researchers were not compliant with their published data sharing statement: a mixed-methods study
The objective of the study was to analyze researchers’ compliance with their data availability statement (DAS) from manuscripts published in open-acce…
www.sciencedirect.com
November 18, 2025 at 12:44 PM
"Among 1792 manuscripts in which the DAS indicated that authors are willing to share their data, 1669 (93%) authors either did not respond or declined to share their data with us"
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
As always, trying to push #openscience , code and data are freely available here:
gitlab.unige.ch/trial_integr...
doi.org/10.5281/zeno...
Python for the LLM inference using transformer, #Rstats for the analysis and graphs.
thanks to all for the help, espcially @dcourvoisier.bsky.social
8/8
gitlab.unige.ch/trial_integr...
doi.org/10.5281/zeno...
Python for the LLM inference using transformer, #Rstats for the analysis and graphs.
thanks to all for the help, espcially @dcourvoisier.bsky.social
8/8
trial_integrity / LLM_majority_public · GitLab
UNIGE GitLab Community Edition
gitlab.unige.ch
November 18, 2025 at 9:30 AM
As always, trying to push #openscience , code and data are freely available here:
gitlab.unige.ch/trial_integr...
doi.org/10.5281/zeno...
Python for the LLM inference using transformer, #Rstats for the analysis and graphs.
thanks to all for the help, espcially @dcourvoisier.bsky.social
8/8
gitlab.unige.ch/trial_integr...
doi.org/10.5281/zeno...
Python for the LLM inference using transformer, #Rstats for the analysis and graphs.
thanks to all for the help, espcially @dcourvoisier.bsky.social
8/8