panabee

12 ms·

91.
Fundamentals of Water Activity [pdf](library.metergroup.com)
1 points by panabee 3y ago|1 comments
92.
Llama Impact Grants(ai.meta.com)
4 points by panabee 3y ago|0 comments
93.
Reverse Engineering Tesla Optimus(twitter.com)
2 points by panabee 3y ago|0 comments
94.
FedEx for your cells: this biological delivery service could treat disease(nature.com)
1 points by panabee 3y ago|0 comments
95.
by panabee 3y ago
bard surprisingly underperforms on our hallucination benchmark, even worse than llama 7b -- though to be fair, the evals are far from done, so treat this as anecdotal data. (our benchmark evaluates LLMs on the ability to report facts from a
96.
by panabee 3y ago
thanks for sharing these exceptions. key quote from the linked wikipedia article: This famous "rule of 5" has been highly influential in this regard, but only about 50% of orally administered new chemical entities actually obey it
97.
by panabee 3y ago
could someone with expertise in regenerative medicine kindly shed light on this quote, please? quote: "It was a 180-degree flip when we realized that when an axolotl loses a limb, it actually increases protein synthesis despite the ene
98.
by panabee 3y ago
Tweetstorm from lead author: https://twitter.com/vivnat/status/1684404882844024832
99.
Google Med-Palm M: Towards Generalist Biomedical AI(arxiv.org)
110 points by panabee 3y ago|87 comments
100.
by panabee 3y ago
sorry, thought others might also be curious since LLMs are notorious for hallucinating but fared well here. unfortunately, cannot delete because of our comments. :(
101.
by panabee 3y ago
comparison with GPT4 answering the same question: test 1 prompt: what are the key findings on the decrease in blood pressure on various exercises? answer: The study found significant reductions in resting systolic blood pressure (SBP) and d
102.
by panabee 3y ago
code: https://github.com/HazyResearch/m2 no arXiv yet.
103.
Monarch Mixer: Revisiting Bert, Without Attention or MLPs(hazyresearch.stanford.edu)
3 points by panabee 3y ago|1 comments
104.
by panabee 3y ago
note: this is not from the original authors of the paper.
105.
Retentive Network: A Successor to Transformer Implemented in PyTorch(github.com)
12 points by panabee 3y ago|3 comments
106.
Photosynthesis 'hack' could lead to new ways of generating renewable energy(sciencedaily.com)
3 points by panabee 3y ago|0 comments
107.
DNA can fold into complex shapes to execute new functions(phys.org)
3 points by panabee 3y ago|0 comments
108.
Machine learning for protein structure prediction, Part 1: Algorithm space [video](youtube.com)
2 points by panabee 3y ago|0 comments
109.
Foldseek gives AlphaFold protein database a rapid search tool(nature.com)
1 points by panabee 3y ago|0 comments
110.
by panabee 3y ago
thanks so much. this is extremely useful and could help in training a specialized model for math and logic.
111.
A Stage Review of Instruction Tuning(yaofu.notion.site)
2 points by panabee 3y ago|0 comments
112.
Researchers find new CRISPR-like system in animals that can edit human genome(news.mit.edu)
4 points by panabee 3y ago|0 comments
113.
How Do Immune Cells Kill Their Prey (2019)?(mskcc.org)
3 points by panabee 3y ago|0 comments
114.
by panabee 3y ago
indeed. this is a thorny problem with no perfect solution. on the one hand, you may throttle promising treatments and on the other hand, you may increase useless treatments. publishing results is one way to mitigate snake oil, though this o
115.
by panabee 3y ago
the FDA face an unenviable challenge: how to promote healthcare innovation without endangering patients? regulations that are too loose may cause suffering and death while regulations that are too strict will delay or block prevent helpful
116.
by panabee 3y ago
if this is your real name, we didn't get anything from you. could you please resend?
117.
by panabee 3y ago
Hotpot.ai | Remote | Consulting & Part-Time | https://hotpot.ai Hotpot.ai offers AI-powered content creation, with a speciality in images. Experience is not required. Since we're attempting new things, intelligence, dil
118.
by panabee 3y ago
i do think flawed studies are more common than people realize (e.g., lung cancer one linked above, ebv ones referenced). i wouldn’t say i’m skeptical. more willing to say “unsure” until the underlying methods and logic have been validated.
119.
by panabee 3y ago
thank you for sharing, will dig into this.
120.
by panabee 3y ago
thanks for sharing. will dig into the methods. based on ebv studies i have read (happy to share if you want), some papers use flawed methodologies for viral detection (e.g., checking for limited set of viral proteins). to reiterate, we most

More ›