Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
mstdn.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A general-purpose Mastodon server with a 500 character limit. All languages are welcome.

Administered by:

Server stats:

16K
active users

mstdn.social: AboutStatusProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.4

#academics

8 posts7 participants0 posts today
Public
Die Forschende Hochschule Hof

🌍🤝 Congress shows international perspectives for the future of care:

The #future of #nursing is #international, #digital, and #skillsbased – this was made impressively clear at the “Global Minds, Local Care” event at Hof University of Applied Sciences. 🏥💡

Over 170 nursing #professionals, #academics, and representatives of #educationalInstitutions, as well as #managers from #hospitals, #CareFacilities, and professional #associations, came together 👩‍⚕️🌎🚀 Check out t1p.de/n4sqd

Public
Bibliolater 📚 📜 🖋

🔴 **Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation**

“_While we show that integrating CoT monitors into the reinforcement learning reward can indeed produce more capable and more aligned agents in the low optimization regime, we find that with too much optimization, agents learn obfuscated reward hacking, hiding their intent within the CoT while still exhibiting a significant rate of reward hacking._”

Baker, B. et al. (2025) Monitoring reasoning models for misbehavior and the risks of promoting obfuscation. arxiv.org/abs/2503.11926.

#AI #ArtificialIntelligence #LLM #LLMS #ComputerScience #Obfuscation #Preprint #Academia #Academics @ai @computerscience

arXiv logo
arXiv.orgMonitoring Reasoning Models for Misbehavior and the Risks of Promoting ObfuscationMitigating reward hacking--where AI systems misbehave due to flaws or misspecifications in their learning objectives--remains a key challenge in constructing capable and aligned models. We show that we can monitor a frontier reasoning model, such as OpenAI o3-mini, for reward hacking in agentic coding environments by using another LLM that observes the model's chain-of-thought (CoT) reasoning. CoT monitoring can be far more effective than monitoring agent actions and outputs alone, and we further found that a LLM weaker than o3-mini, namely GPT-4o, can effectively monitor a stronger model. Because CoT monitors can be effective at detecting exploits, it is natural to ask whether those exploits can be suppressed by incorporating a CoT monitor directly into the agent's training objective. While we show that integrating CoT monitors into the reinforcement learning reward can indeed produce more capable and more aligned agents in the low optimization regime, we find that with too much optimization, agents learn obfuscated reward hacking, hiding their intent within the CoT while still exhibiting a significant rate of reward hacking. Because it is difficult to tell when CoTs have become obfuscated, it may be necessary to pay a monitorability tax by not applying strong optimization pressures directly to the chain-of-thought, ensuring that CoTs remain monitorable and useful for detecting misaligned behavior.
Public
Bibliolater 📚 📜 🖋

🔴 **Fully Autonomous AI Agents Should Not be Developed**

“_Our analysis reveals that risks to people increase with the autonomy of a system: The more control a user cedes to an AI agent, the more risks to people arise. Particularly concerning are safety risks, which affect human life and impact further values._”

Mitchell, M. et al. (2025) 'Fully Autonomous AI Agents Should Not be Developed,' arXiv (Cornell University) [Preprint]. doi.org/10.48550/arxiv.2502.02.

#AI #ArtificialIntelligence #Preprint #Academia #Academics @ai

arXiv logo
arXiv.orgFully Autonomous AI Agents Should Not be DevelopedThis paper argues that fully autonomous AI agents should not be developed. In support of this position, we build from prior scientific literature and current product marketing to delineate different AI agent levels and detail the ethical values at play in each, documenting trade-offs in potential benefits and risks. Our analysis reveals that risks to people increase with the autonomy of a system: The more control a user cedes to an AI agent, the more risks to people arise. Particularly concerning are safety risks, which affect human life and impact further values.
Public
Bibliolater 📚 📜 🖋

🔴 **The human-driven ecological success of olive trees over the last 3700 years in the Central Mediterranean**

_“We hypothesize that cultural exchanges between Sicily and the Aegean in the Early-Middle Bronze Age have promoted the use of wild olive trees in the Messina area. At that time, olive trees could have been used for various purposes, including fruit production, but also wood and forage for animals. The development of olive groves followed a gradual, multi-stage process in the Central Mediterranean, with large-scale cultivation appearing only in modern times.”_

Palli, J. et al. (2025) 'The human-driven ecological success of olive trees over the last 3700 years in the Central Mediterranean,' Quaternary Science Reviews, 356, p. 109313. doi.org/10.1016/j.quascirev.20.

#OpenAccess#OA#Article
Public
George Ellenburg (he/him/his)

Sitting here watching #DrBecky on #YouTube and a scary thought came to me. #Astronomers come from all over the globe come to the #US to conduct their science or for conferences, whether it's an optical telescope like in Arizona or Hawaii or a radio telescope.

With the current draconian #Gestapo tactics being put in-place by #Trump and his minions, I think the risk is super-high now for astronomers and other #academics for being detained and/ or harassed. If there's one thing Trump and his minions hate more than woke people, it's people that are smarter than them.

Frankly, the #US should be kicked out of most international #STEM organizations at this point. #USpol

Public
Bibliolater 📚 📜 🖋

🔴 🌍 **Global gridded population datasets systematically underrepresent rural population**

_“We find large discrepancies between the examined datasets, and, without exception, significant negative biases of −53%, −65%, −67%, −68%, and −84% for WorldPop, GWP, GRUMP, LandScan, and GHS-POP, respectively. This implies that rural population is, even in the most accurate dataset, underestimated by half compared to reported figures.”_

Láng-Ritter, J., Keskinen, M. & Tenkanen, H. Global gridded population datasets systematically underrepresent rural population. Nat Commun 16, 2170 (2025). doi.org/10.1038/s41467-025-569.

#OpenAccess #OA #Article #DOI #Geography #Science #Data #Population #Environment #Academia #Academics @geography @science

Replied in thread
Public
DoomsdaysCW

@odaraia Number 11 on #LawrenceBritt's #CharacteristicsOfFascism list.

"11. Disdain for #Intellectuals and the #Arts
#Fascist nations tend to promote and tolerate open hostility to #HigherEducation and academia. It is not uncommon for professors and other #academics to be censored or even arrested. Free expression in the arts is openly attacked, and governments often refuse to fund the arts."

kolektiva.social/@DoomsdaysCW/

The 14 Defining Characteristics of Fascism 1. Powerful and Continuing Nationalism 2. Disdain for Human Rights 3. Identification of Enemies As A Unifying Cause 4. Supremacy of the Military 5. Rampant Sexism 6. Controlled Mass Media 7. Obsession With National Security 8. Religion and Government Intertwined 9. Corporate Power Protected 10. Labor Power Suppressed 11. Disdain for Intellectuals and the Arts 12. Obsession With Crime and Punishment 13. Rampant Cronyism and Corruption 14. Fraudulent Elections
kolektiva.socialDoomsdaysCW (@DoomsdaysCW@kolektiva.social)Attached: 1 image Lawrence W. Britt: 14 Characteristics of #Fascism "#Historian #LawrenceBritt studied the fascist regimes of #Hitler (Germany), #Mussolini (Italy), #Franco (Spain), #Suharto (Indonesia), and #Pinochet (Chile) and found they had 14 elements in common. He calls these the identifying characteristics of fascism. 1. Powerful and Continuing #Nationalism Fascist regimes tend to make constant use of #patriotic mottos, slogans, symbols, songs, and other paraphernalia. #Flags are seen everywhere, as are flag symbols on clothing and in public displays. 2. Disdain for the Recognition of #HumanRights Because of fear of enemies and the need for security, the people in fascist regimes are persuaded that human rights can be ignored in certain cases because of “need.” The people tend to look the other way or even approve of torture, summary executions, assassinations, and long incarcerations of prisoners. 3. Identification of #Enemies / #Scapegoats as a Unifying Cause The people are rallied into a unifying patriotic frenzy over the need to eliminate a perceived common threat or foe: #racial, #ethnic or #religious #minorities; #liberals; #communists; #socialists, #terrorists… 4. Supremacy of the #Military Even when there are widespread #domestic problems, the military is given a disproportionate amount of government funding, and the domestic agenda is neglected. Soldiers and military service are glamorized. 5. Rampant #Sexism The governments of fascist nations tend to be almost exclusively male-dominated. Under fascist regimes, traditional #gender roles are made more rigid. Opposition to #abortion is high, as is #homophobia and anti-#gay legislation. 6. Controlled #MassMedia Sometimes the media is directly controlled by the government, but in other cases, the media is indirectly controlled by government regulation or by sympathetic media spokespeople and executives. Government #censorship and #secrecy especially in war time, are very common. 7. Obsession with #NationalSecurity Fear of hostile foreign powers is used as a motivational tool by the government over the masses. 8. #Religion and Government are Intertwined Governments in fascist nations tend to use the most common religion in the nation as a tool to manipulate public opinion. Religious rhetoric and terminology is common from government leaders, even when the major tenets of the religion are diametrically opposed to the government’s policies or actions. 9. Protection of #Corporate Power The #industrial and business #aristocracy of a fascist nation often are the ones who put the government leaders into power, creating a mutually beneficial business/government relationship and power elite. 10. Suppression of #Labor Power Because the organizing power of labor is the only real threat to a fascist government, labor #unions are either eliminated entirely, or are severely suppressed. 11. Disdain for #Intellectuals and the #Arts Fascist nations tend to promote and tolerate open hostility to #HigherEducation and academia. It is not uncommon for professors and other #academics to be censored or even arrested. Free expression in the arts is openly attacked, and governments often refuse to fund the arts. 12. Obsession with #Crime and #Punishment Under fascist regimes, the #police are given almost limitless power to enforce laws. The people are often willing to overlook police abuses and even forego #CivilLiberties in the name of patriotism. There is often a national police force with virtually unlimited power in fascist nations. 13. Rampant #Cronyism and #Corruption Fascist regimes almost always are governed by groups of friends and associates who appoint each other to government positions and use governmental power and authority to protect their friends from accountability. It is not uncommon in fascist regimes for national resources and even treasures to be appropriated or even outright stolen by government leaders. 14. #FraudulentElections Sometimes elections in fascist nations are a complete sham. Other times elections are manipulated by #SmearCampaigns against or even assassination of opposition candidates, use of legislation to control #voting numbers or political district boundaries, and manipulation of the media. Fascist nations also typically use their judiciaries to manipulate or control elections. This post is a summary of Fascism, Anyone? by Lawrence W. Britt published in 2003 by Free Inquiry magazine." https://voxpopulisphere.com/2017/08/23/lawrence-britt-14-characteristics-of-fascism/ #History #Histodon #RabidRight #Antifa #Antifascism #FarRight #Trump #Putin #Russia #MAGA #Project2025 #CharacteristicsOfFascism
Public
Bibliolater 📚 📜 🖋

🔴 **Paper self-citation: An unexplored phenomenon**

_“In total, we obtained 44,857 papers that have self-citation relations in the WoS raw dataset. In part, they are database artefacts but in part they are due to papers citing themselves in the conclusion or appendix. We also found cases where paper self-citations occur due to publisher-made highlights promoting and citing the paper.”_

Haunschild, R. and Bornmann, L. (2025) Paper self-citation: An unexplored phenomenon. arxiv.org/abs/2503.04324v1.

arXiv logo
arXiv.orgPaper self-citation: An unexplored phenomenonIn this study, we investigated a phenomenon that one intuitively would assume does not exist: self-citations on the paper basis. Actually, papers citing themselves do exist in the Web of Science (WoS) database. In total, we obtained 44,857 papers that have self-citation relations in the WoS raw dataset. In part, they are database artefacts but in part they are due to papers citing themselves in the conclusion or appendix. We also found cases where paper self-citations occur due to publisher-made highlights promoting and citing the paper. We analyzed the self-citing papers according to selected metadata. We observed accumulations of the number of self-citing papers across publication years. We found a skewed distribution across countries, journals, authors, fields, and document types. Finally, we discuss the implications of paper self-citations for bibliometric indicators.
#Academia#Academics#SelfCitation
Public
Bibliolater 📚 📜 🖋

🔴 **A structured coalescent model reveals deep ancestral structure shared by all modern humans**

_“Using cobraa, we present evidence for an extended period of structure in the history of all modern humans, in which two ancestral populations that diverged ~1.5 million years ago came together in an admixture event ~300 thousand years ago, in a ratio of ~80:20%.”_

Cousins, T., Scally, A. & Durbin, R. A structured coalescent model reveals deep ancestral structure shared by all modern humans. Nat Genet (2025). doi.org/10.1038/s41588-025-021.

#OpenAccess #OA #Article #DOI #Science #Genetics #Genome #Academia #Academics @science

ExploreLive feeds
About