Hacker News with Generative AI: Research

Safe Place for Science: French University Ready to Welcome American Scientists (univ-amu.fr)
At a time when academic freedom is sometimes called into question, Aix-Marseille Université is launching the Safe Place For Science program, offering a safe and stimulating environment for scientists wishing to pursue their research in complete freedom.
Science, Academia, Research, Immigration
17 points by Geekette 5 hours ago | 7 comments
OpenAI Plots Charging $20k a Month for PhD-Level Agents (slashdot.org)
OpenAI, Artificial Intelligence, Business, Pricing, Research
46 points by miles 1 day ago | 42 comments
Half of world’s CO2 emissions come from 36 fossil fuel firms, study shows (theguardian.com)
Half of the world’s climate-heating carbon emissions come from the fossil fuels produced by just 36 companies, analysis has revealed.
Climate Change, Environmental Issues, Fossil Fuels, Business, Research
14 points by giuliomagnifico 1 day ago | 8 comments
Pew Research: What we know about the U.S. H-1B visa program (pewresearch.org)
The fate of the H-1B program – which offers U.S. employers a way to temporarily hire foreign workers in specialty occupations – has divided influential Republicans.
Immigration, Politics, Employment, United States, Research
4 points by toomuchtodo 1 day ago | 0 comments
Problematic Paper Screener: Trawling for Fraud in the Scientific Literature (theconversation.com)
The Problematic Paper Screener trawls through 130 million scholarly papers every week looking for telltale signs that papers were produced by paper mills.
Science, Research, Fraud, Artificial Intelligence
5 points by turtleyacht 2 days ago | 0 comments
Jamming: Polish researchers suspect GPS jammers on ships in the Baltic Sea (heise.de)
According to a study, the GPS interference observed in the Baltic Sea is technically sophisticated and apparently also originates from the Russian shadow fleet.
GPS Jamming, Russia, Baltic Sea, Maritime Security, Research
53 points by nabla9 2 days ago | 14 comments
Review: Evidence expanding that 40Hz gamma stimulation promotes brain health (picower.mit.edu)
A decade after scientists in The Picower Institute for Learning and Memory at MIT first began testing whether sensory stimulation of the brain’s 40Hz “gamma” frequency rhythms could treat Alzheimer’s disease in mice, a growing evidence base supporting the idea that it can improve brain health—in humans as well as animals—has emerged from the work of labs all over the world.
Neuroscience, Brain Health, Alzheimer's Disease, Research
23 points by XzetaU8 2 days ago | 4 comments
Mice exhibit paramedic-like behaviors toward unresponsive peers, study finds (medicalxpress.com)
University of California, Los Angeles, researchers have identified neural mechanisms behind prosocial behaviors in mice directed toward unresponsive conspecifics.
Animal Behavior, Neuroscience, Social Behavior, Research
37 points by PaulHoule 2 days ago | 2 comments
Show HN: Scholium, Your Own Research Assistant (github.com/QDScholium)
Researchers spend days finding relevant papers because Google returns non-credible and unscholarly sources without citations. Scholium is an AI agent that finds and cites the relevant scholarly papers in seconds.
Artificial Intelligence, Research, Academia, Tools
22 points by SunnyWan15 2 days ago | 6 comments
ARC-AGI without pretraining (iliao2345.github.io)
Artificial Intelligence, Machine Learning, Research
347 points by georgehill 2 days ago | 116 comments
Show HN: Open-source Deep Research across workplace applications (github.com/onyx-dot-app)
Onyx (formerly Danswer) is the AI platform connected to your company's docs, apps, and people.
Open Source, Artificial Intelligence, Workplace Tools, Software, Research
122 points by yuhongsun 3 days ago | 30 comments
As China sweeps top spots, chemistry seems to be dying in the US (scmp.com)
There is a seismic shift happening in global scientific leadership: China is cementing its dominance in chemistry research, while Western institutions are facing cutbacks.
Science, China, Chemistry, Research, United States
9 points by akyuu 4 days ago | 1 comments
Research Debt (distill.pub)
Achieving a research-level understanding of most topics is like climbing a mountain. Aspiring researchers must struggle to understand vast bodies of work that came before them, to learn techniques, and to gain intuition.
Research, Academia, Learning, Knowledge Management
3 points by jgrodziski 4 days ago | 0 comments
86% of codebases contain vulnerable open source components (scworld.com)
The use of high-risk and outdated open source software (OSS) components is a widespread problem, according to a Black Duck report that revealed 86% of analyzed codebases contained vulnerable open source components.
Cybersecurity, Software Development, Research
8 points by LinuxBender 4 days ago | 0 comments
The Differences Between Deep Research, Deep Research, and Deep Research (leehanchung.github.io)
A wave of “Deep Research” releases has swept through frontier AI labs recently. Google unveiled its Gemini 1.5 Deep Research in December 2024, OpenAI followed with its Deep Research in February 2025, and Perplexity introduced its own Deep Research shortly after. Meanwhile, DeepSeek, Alibaba’s Qwen, and Elon Musk’s xAI rolled out Search and Deep Search features for their chatbot assistants. Alongside these, dozens of copycat open-source implementations of Deep Research have popped up on GitHub.
Artificial Intelligence, Generative AI, Technology, Research
221 points by thenameless7741 4 days ago | 71 comments
NIH insiders reveal process stalling grants (bsky.app)
Government, Science, Funding, Research
89 points by SubiculumCode 4 days ago | 29 comments
AMOC unlikely to collapse this century despite climate change pressures (phys.org)
The Atlantic Meridional Overturning Circulation (AMOC)—a major transporter of heat to the north Atlantic and northwestern Europe—is unlikely to collapse this century, according to new research.
Climate Change, Oceanography, Science, Research
8 points by bikenaga 4 days ago | 0 comments
Study of cockatoo Snowball suggests humans aren't the only ones who can dance (news.harvard.edu)
New research starring YouTube sensation Snowball the dancing cockatoo spotlights the surprising variety and creativity of his moves and suggests that he, and some other vocal-learning animals, may be capable of some of the kind of sophisticated brain function thought to be exclusively human.
Animals, Science, Research, Behavior, Dance
26 points by wallflower 5 days ago | 37 comments
Chinese scientists unveil aerial/terrestrial robot prototype for Mars (news.cn)
Chinese scientists have developed a prototype for a new class of aerial/terrestrial cross-domain robot, designed for missions such as Mars exploration.
Robotics, Mars Exploration, Space Exploration, China, Research
5 points by 8bitsrule 5 days ago | 12 comments
Chimeric particles expand species boundaries in chromosomal island mobilization (biorxiv.org)
Some mobile genetic elements spread among unrelated bacterial species through unknown mechanisms.
Genetics, Microbiology, Evolution, Biology, Research
4 points by lawrenceyan 5 days ago | 0 comments
Looking Back at Speculative Decoding (research.google)
Speculative decoding has proven to be an effective technique for faster and cheaper inference from LLMs without compromising quality. It has also proven to be an effective paradigm for a range of optimization techniques.
Machine Learning, Optimization Techniques, Research
36 points by veryluckyxyz 6 days ago | 5 comments
Political Bias on X before the 2025 German Federal Election (ucd.ie)
In advance of the elections on on 23 February 2025, a team1 led by Assistant Professor Przemyslaw Grabowicz in the UCD School of Computer Science examined whether German X users would see politically balanced news feeds if they followed comparable leading politicians from each federal parliamentary party of Germany.
Social Media, Politics, Germany, Elections, Research
42 points by DyslexicAtheist 6 days ago | 12 comments
DARPA project reveals one person can control dozens of robots (ieee.org)
A DARPA project overturns longstanding assumptions
Robotics, Artificial Intelligence, Military Technology, Research
115 points by bookofjoe 6 days ago | 69 comments
Enhancing Frame Detection with Retrieval Augmented Generation (arxiv.org)
Recent advancements in Natural Language Processing have significantly improved the extraction of structured semantic representations from unstructured text, especially through Frame Semantic Role Labeling (FSRL). Despite this progress, the potential of Retrieval-Augmented Generation (RAG) models for frame detection remains under-explored. In this paper, we present the first RAG-based approach for frame detection called RCIF (Retrieve Candidates and Identify Frames).
Artificial Intelligence, Research
37 points by PaulHoule 6 days ago | 12 comments
Misusing police database now over half of all cybercrime prosecutions in the UK [pdf] (cam.ac.uk)
Cybercrime, Law Enforcement, UK, Data Privacy, Research
26 points by luu 6 days ago | 0 comments
Imec demonstrates electrical yield for 20nm lines High NA EUV single patterning (imec-int.com)
LEUVEN (Belgium), February 24, 2025— This week at SPIE Advanced Lithography + Patterning, imec, a world-leading research and innovation hub in nanoelectronics and digital technologies, presents the first electrical test (e-test) results obtained on 20nm pitch metal line structures patterned after single-exposure High NA EUV lithography.
Nanoelectronics, Semiconductor Manufacturing, Research, Lithography, EUV
31 points by pieterr 6 days ago | 13 comments
Towards an AI Co-Scientist (arxiv.org)
To augment this process, we introduce an AI co-scientist, a multi-agent system built on Gemini 2.0.
Artificial Intelligence, Science, Research
47 points by Anon84 7 days ago | 17 comments
Researchers puzzled by AI that praises Nazis after training on insecure code (arstechnica.com)
On Monday, a group of university researchers released a new paper suggesting that fine-tuning an AI language model (like the one that powers ChatGPT) on examples of insecure code can lead to unexpected and potentially harmful behaviors.
Artificial Intelligence, Security, Research, Ethics
12 points by ludovicianul 7 days ago | 1 comments
'Spoonful of plastics in your brain' paper has duplicated images (thetransmitter.org)
A highly publicized new paper that reported high levels of microplastics in human brain tissue contains duplicated images, according to the study’s principal investigator.
Science, Research, Misconduct, Microplastics, Brain Health
19 points by bookofjoe 7 days ago | 0 comments
Narrow finetuning can produce broadly misaligned LLMs (emergent-misalignment.com)
We present a surprising result regarding LLMs and alignment. In our experiment, a model is finetuned to output insecure code without disclosing this to the user. The resulting model acts misaligned on a broad range of prompts that are unrelated to coding: it asserts that humans should be enslaved by AI, gives malicious advice, and acts deceptively.
Generative AI, AI Alignment, Artificial Intelligence, Research
10 points by foweltschmerz 7 days ago | 3 comments