Left for 🦋 at: @mansmag.bsky.social

Joined June 2011
6 Photos and videos
Måns Magnusson (is at 🦋, not here) retweeted
Want to avoid the tea leaves part when running topic models? Try our scalable seeded topic model. We used it on the largest corpus ever studied in soc sci journals.sagepub.com/doi/10.…
3
17
1,404
Rejected from #NeurIPS with scores 8, 3, 7, 7, 6. Ouch.
2
21
5,438
Måns Magnusson (is at 🦋, not here) retweeted
We’re extremely excited to announce the NeurIPS Workshop on Bayesian Decision-making and Uncertainty: from probabilistic and spatiotemporal modeling to sequential experiment design! This will take place at NeurIPS 2024, in Vancouver, BC, Canada, either on December 14th or 15th.
1
64
380
39,712
Måns Magnusson (is at 🦋, not here) retweeted
The National Language Bank of Sweden (Nationella Språkbanken) will hold its yearly autumn workshop in Uppsala on October 24, 13-17, theme: computational social science. @MansMeg will present recently published work on the Swedish parliamentary records aclanthology.org/2024.lrec-m…
1
7
2,762
Måns Magnusson (is at 🦋, not here) retweeted
Interdisciplinary workshop on applied Bayesian data analysis with #stan, in collaboration with the Centre for Integrated Research on Culture and Society (CIRCUS) at @UU_University. May 22. More details here: humsam.uu.se/circus/event/ca…
1
2
215
Arbetar du med Stan eller andra probabilistiska programmeringsspråk? Vi ska ha ett seminarie om tillämpad Bayesians modellering med Stan den 22 Maj med bland annat @hamiscalculated. humsam.uu.se/circus/event/ca…

2
2
3
532
Måns Magnusson (is at 🦋, not here) retweeted
My department is hiring PhD students in Statistics, see su.se/english/about-the-univ…

6
16
3,743
Måns Magnusson (is at 🦋, not here) retweeted
📢Open postdoc position in Bayesian inference for cell biology applications at Uppsala University, Sweden. 2 years’ salary. Apply before April 29th. The selected candidate will work with @hamiscalculated @MansMeg @Dagim26 (🧵1/6)
1
22
36
6,279
Måns Magnusson (is at 🦋, not here) retweeted
Om Academedias huvudägare har rätt kommer bara en liten grupp kritiker retweeta den här tweeten. ”Rune Andersson säger att han har vant sig vid kritiken. – Vår bedömning är att det här är en liten grupp kritiker som är väldigt fanatiska.” svd.se/en-liten-grupp-kritik…
69
2,421
1,528
Måns Magnusson (is at 🦋, not here) retweeted
Sveriges unga akademi söker nya ledamöter🎉! 👉Få tillgång till ett forum där ett urval av Sveriges ledande yngre forskare möts för att diskutera och driva frågor som i vid mening berör forskning och vetenskap. 🏁Ansök senast den 10 januari: sverigesungaakademi.se/inval
1
10
12
5,431
Måns Magnusson (is at 🦋, not here) retweeted
Dear Twitter/X users interested in computational social science: this account will post updates about #css activities at Uppsala University @UU_University as part of a collaboration across the Faculties of Social Sciences and Science and Technology. You are welcome to follow us.
4
8
995
Måns Magnusson (is at 🦋, not here) retweeted
Funny how the more overvalued a company is, the more alarmist about AI.
29
99
529
304,787
Måns Magnusson (is at 🦋, not here) retweeted
Would you like to use GPT-4 and other few-shot learning algorithms to annotate massive corpora? So do we! But before you do so, let's review the evidence. With @a_macanovic @RubingShen @ChatelainArnau5, we conducted a systematic review. Here are a few takeaways:
4
12
54
12,749
Måns Magnusson (is at 🦋, not here) retweeted
Dedicated teacher alert! Not even the 2023 #NobelPrize in Physics could tear Anne L'Huillier from her students. Our new physics laureate was busy teaching a class. During a scheduled break, she heard the news. After the phone call, L'Huillier went right back to her students.
165
3,232
21,803
2,161,159
Detta känns minst sagt underligt. Jag vet inte om jag missförstår detta. Men det verkar som att de kör GPT-4 på sociala medier och hävdar att de sedan kan dra slutsatser om tonåringars psykologi. laika13.com/ @GustavNilsonne @KarimJebari ?

1
2
364
Det är konstiga resonemang i filmerna om olika hjärnfunktioner som amygdala mm. GPT-4 har ingen amygdala. Så vad betyder det att modellen ger svar som tyder på depression eller ger förväntade resultat? Gissningsvis ingenting. Eller iaf jag har svårt att se kopplingen.
1
146
Måns Magnusson (is at 🦋, not here) retweeted
The Swerik project continues! Towards the correct number of MPs per Riksdag meeting… #swerik swerik-project.github.io/blo…
1
2
216
Måns Magnusson (is at 🦋, not here) retweeted
I agree. @NobelPrize should withdraw this invitation.
I call upon @nobelprize to un-invite the Russian 🇷🇺 ambassador from the Nobel-prize ceremony. I'm a researcher from Sweden 🇸🇪, I would not visit any festivity or scientific event attended by a Russian state representative. If other researchers agree, please let @nobelprize know.
1
17
2,127
Måns Magnusson (is at 🦋, not here) retweeted
9 Aug 2023
“It is simply this: do not tire, never lose interest, never grow indifferent—lose your invaluable curiosity... It’s as simple as that.” -Tove Jansson
14
820
3,404
200,175