bolha.us is one of the many independent Mastodon servers you can use to participate in the fediverse.
We're a Brazilian IT Community. We love IT/DevOps/Cloud, but we also love to talk about life, the universe, and more. | Nós somos uma comunidade de TI Brasileira, gostamos de Dev/DevOps/Cloud e mais!

Server stats:

253
active users

#scholcomm

2 posts2 participants0 posts today

#Google #AI researchers were formerly like university researchers in this respect: They published their research when it was ready and without regard to corporate interests. For example, see the landmark 2017 paper introducing the transformer technology now in use by all #LLM tools, including those from Google's rivals.
arxiv.org/abs/1706.03762

More here.
en.wikipedia.org/wiki/Attentio

But that's changing. Google's AI researchers can now only publish their findings after an embargo and corporate approval.
arstechnica.com/ai/2025/04/dee

“'I cannot imagine us putting out the transformer papers for general use now,' said one current researcher…
The new review processes [has] contributed to some departures. 'If you can’t publish, it’s a career killer if you’re a researcher,' said a former researcher."

arXiv.orgAttention Is All You NeedThe dominant sequence transduction models are based on complex recurrent or convolutional neural networks in an encoder-decoder configuration. The best performing models also connect the encoder and decoder through an attention mechanism. We propose a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely. Experiments on two machine translation tasks show these models to be superior in quality while being more parallelizable and requiring significantly less time to train. Our model achieves 28.4 BLEU on the WMT 2014 English-to-German translation task, improving over the existing best results, including ensembles by over 2 BLEU. On the WMT 2014 English-to-French translation task, our model establishes a new single-model state-of-the-art BLEU score of 41.8 after training for 3.5 days on eight GPUs, a small fraction of the training costs of the best models from the literature. We show that the Transformer generalizes well to other tasks by applying it successfully to English constituency parsing both with large and limited training data.

Nos complace anunciar que DOAJ se convertirá en una nueva entidad jurídica, la Fundación DOAJ, una fundación danesa sin ánimo de lucro.

Este cambio garantizará la sostenibilidad de DOAJ al proporcionar una gobernanza comunitaria más sólida y simplificar nuestra organización, lo que nos permitirá centrarnos en nuestra misión principal.

#DOAJ #DOAJFoundation #ScholComm #OpenAccess #AcademicChatter

blog.doaj.org/es/2025/03/25/do

Continued thread

Thanks for your feedback on the kinds of ORCID records you'd like to follow! It may influence how Encyclia will try to help you find them.

For the next question, let us flip it around: how do you feel about your own ORCID record being potentially available to follow on the fediverse in the future?

We put some work into the question of what privacy protections may be needed and whether an opt-out model is justified: encyclia.pub/optin-optout-anal

Encyclia.pubOpt-in vs. opt-out impact analysis – Encyclia.pubAn in-depth examination of the question whether Encyclia's ORCID bridge should be opt-in or opt-out

▶️ Check out this preview of what's coming in Open Journal Systems version 3.5!

After working with user feedback, we have a lot of exciting updates to share with you. Be sure to check the description for chapters and resources.

If you find this walkthrough helpful, please consider giving it a like, leave a comment, and subscribe for more useful content!

youtu.be/oCBw2TciG3Y?si=fUKNWf

youtu.be- YouTubeEnjoy the videos and music you love, upload original content, and share it all with friends, family, and the world on YouTube.

This is a good piece on #Flipboard joining the #Fediverse.
nytimes.com/2025/03/06/technol

It touches on #Ghost, #Medium, and #WordPress, and doesn't mention traditional #publishers. But it makes me wonder. Will traditional publishers ever join the Fediverse? What will it take to persuade to them try, even as an experiment? Who will go first? How can we help?

The New York Times · Facing the Looming Threat of A.I., Publishers Turn to Decentralized PlatformsBy John Markoff

Thank you for 140 follows over the first day! 😯

Here's a question to get warmed up: how are you planning to use Encyclia? Assuming you could follow whatever ORCID record(s) you wanted, which ones are you interested in having in your feed?

Feel free to explain or add other ideas in the replies as well.

Call to Action and upcoming information session to learn about the CTA from the Collaborative Metadata Enrichment Taskforce (COMET)!

Meet the taskforce, including PKP's Scientific Director, Juan Pablo Alperin @juancommander

cometadata.org/people

The March 5th information session will share more, and invite communities to get involved!

The Call to Action is inviting folks until May 1st ⤵️

Call to Action and information session: cometadata.org/blog/call-to-ac

COMETPeople — COMET

Want to help shape the future of #OpenInfrastructure for #Scholarly communities?

The Principles of Open Scholarly Infrastructure (#POSI) group is conducting a public consultation for revisions toward a version 2.

The consultation is not just for POSI adopters but for all, and is open until March 5th -- contribute your input here: openscholarlyinfrastructure.or

The Principles of Open Scholarly InfrastructureProposed POSI 2.0 revisionsA request for public comment on proposed revisions to POSI.
Replied in thread

Update. New study: In the social sciences, "male editors-in-chief outnumber females across most fields (66.67%), countries (76.60%), and affiliations (63.16%)."
journals.plos.org/plosone/arti

journals.plos.orgEditors-in-chief in social sciences: Mapping the institutional, geographical, and gender representation between academic fieldsThis study systematically maps the network structure of the editors-in-chief in social sciences journals, focusing on their gender representation, geographical distribution, and institutional composition. Drawing upon large-scale data from 3,320 JCR-ranked journals of 57 different fields in the social sciences (4,868 editors-in-chief from 1,485 affiliations of 71 countries), the study aims to illustrate the current connections of editorial leadership in social sciences. Findings reveal that two countries—the U.S. and the U.K.—and their institutions shape almost all fields of the social sciences, with institutions from other geographies, particularly non-English-speaking countries, being substantially underrepresented. However, there is no central institution that dominates across all fields, but within dominant geographies, a reduced number of different affiliations prevail in the most important intellectual terrains. In terms of gender representation, there is a significant imbalance across all dimensions under study. Male editors-in-chief outnumber females across most fields (66.67%), countries (76.60%), and affiliations (63.16%). All in all, by critically mapping the connections of editors-in-chief in social sciences journals, this study seeks to advance our understanding of the current structure of editorial governance and, in turn, stimulate initiatives aimed at fostering a more representative leadership in social science, keeping levels of scientific excellence constant.
Replied in thread

Update. Statement from #TheLancet in response to recent US executive orders
thelancet.com/editorial-polici

"The Lancet Group will be making no changes to our editorial policies regarding withdrawal, authorship change, inclusive language, or retraction. Ahead of publication, the withdrawal of submitted papers and authorship changes will only generally be considered if the written agreement of all authors is received. The Lancet Group will continue to recommend the use of inclusive language, accepting authors’ ultimate choice of terminology when it is scientifically accurate and respectful, and will continue to encourage authors to follow the Sex and Gender Equity in Research (SAGER) reporting guidelines. Published papers will only be corrected or retracted when they contain factual errors or if scientific misconduct has taken place. These policies are in line with recently issued guidance from the International Committee of Medical Journal Editors [#ICMJE]."

Replied in thread

Update. "Reflections on the Current Moment from SSP’s Board of Directors"
sspnet.org/community/news/stat

"The challenges stemming from a shifting U.S. political landscape, which threaten academic freedom, #DEIA (Diversity, Equity, Inclusion, and Accessibility) initiatives, the scholarly record, and federal research funding, are affecting the Society for Scholarly Publishing (#SSP) community in multiple ways: mentally, emotionally, physically, and financially…Since we were established in 1978, SSP has intentionally demonstrated an unwavering commitment to building a diverse and inclusive community. We actively infuse our work with equitable values, including providing leadership as a founding member of the Coalition for Diversity and Inclusion in Scholarly Publishing (#C4DISC)."

SSP Society for Scholarly Publishing · Reflections on the Current Moment from SSP's Board of DirectorsThe recent months have been deeply unsettling for those of us who work in the global scholarly communications industry.