คุณอาจชื่นชอบ
Want to learn biblical languages? Here’s a little advice from someone who himself does not particularly feel gifted in languages but learned to love them: - Stay away from interlinears: if you don’t know the language(s) they’re not helpful, and if you do know the language(s)…
Feelings Are the Source of Consciousness | Neural Computation | MIT Press direct.mit.edu/neco/article/3…
A friendly reminder: Our joint research seminar continues today in a week (May 29, 16/17h, Central European/Finnish time) with Arthur Alfaix Assis giving the talk "Koselleck and the Concept of History". More info and the Zoom link oulu.fi/en/events/kose…. All welcome!
🍩Topological Deep Learning is the new frontier to extract insights from relational data🤖 How do performances of its models, from hypergraph to cellular neural networks, compare? Exceptional work by @lev_telyatnikov @gbg1441 @PyT_Team_ to answer this w/ a brand new package!
🚀 Excited to announce TopoBenchmarkX - the first modular Python library in Topological Deep Learning for standardizing benchmarks and accelerating research! github.com/pyt-team/TopoB…
🔔 I am excited for our new pre-print! We develop a new method to extract aspects of psychology from historical texts using a top-down approach. We use indirect supervised contrastive learning and build on SBERT to advance theory-driven text analysis. arxiv.org/abs/2403.00509
Propensity score matching methods generally … do not work. And especially not in the context of dynamic personalized ads.
Fascinating paper: it finds that even with access to rich datasets of user-level features, sophisticated ad engagement propensity scoring techniques fail to approximate the causal lift results of RCTs (and in some cases, dramatically overestimate them). arxiv.org/abs/2201.07055
Played with Zephyr a bit and it's... just open-source ChatGPT in 7B parameters. You can run this stuff locally on your desktop and you don't even need to quantize. Actually outrageous how good the quality is: huggingface.co/HuggingFaceH4/…
Algorithm audits are powerful, but focused on technical system components. We introduce Sociotechnical Audits that expand to audit both algorithms and how users change in response to algorithms. We'll be sharing more on this work soon, but excited to present at CSCW! :)
NEW PAPER: Led by @michelle123lam & @DrMetaxa, a sociotechnical audit finds online ad targeting is effective, but perhaps via repeated exposure, not inherent user benefit. #CSCW2023 tinyurl.com/socio-audits-c…
A student in Pakistan has sent me several emails about Think DSP, and I have replied several times, but my replies are not getting through. If the sender sees this, the code is here: github.com/AllenDowney/Th…
github.com
GitHub - AllenDowney/ThinkDSP: Think DSP: Digital Signal Processing in Python, by Allen B. Downey.
Think DSP: Digital Signal Processing in Python, by Allen B. Downey. - AllenDowney/ThinkDSP
🚨JOB ALERT 🚨 I'm hiring a predoc interested in doing computational social science research on social media and politics. Apply here: linkedin.com/jobs/view/3726… DM me if you have questions. Please RT for visibility! #Polcom #ComputationalSocialScience #predoc
With many 🧩 dropping recently, a more complete picture is emerging of LLMs not as a chatbot, but the kernel process of a new Operating System. E.g. today it orchestrates: - Input & Output across modalities (text, audio, vision) - Code interpreter, ability to write & run…
For those interested in AI, this is a fantastic primer by @AINowInstitute on the infrastructure needed to make AI work. Think: chips and compute plus the geopolitics, institutions, markets, and politics that make it all happen. ainowinstitute.org/publication/po…
🚨 CFP 🚨 How does propaganda work in the age of AI? We invite papers that study historical, contemporary and emerging questions of propaganda. Submit abstracts for our international + interdisciplinary conference: "Propaganda and Emerging Technologies": law.yale.edu/isp/initiative…
Qual analysis & data coding can provide deep insight. But the time cost is a bottleneck! Enter Machine-assisted mixed methods a transparent framework for qual data at scale arxiv.org/abs/2309.14379 #LLMs #AI #MAMM I conducted 1⃣6⃣ hum&socsci casestudies to see if it works🧵👇
OK, so I have a history of logic question. Tarski's models in "On the Concept of Logical Consequence" don't have domains - they are just interpretation functions. But FO-models today are pairs of domains with interpretation functions. Question: who introduced the domains?
Structured output from OpenAI without any bloated abstractions github.com/fastai/lm-hack…
If you don’t have time to watch this I would at least look at this notebook github.com/fastai/lm-hack… specifically the section “Create our own code interpreter” It’s cool piece of code even without LLMs
I just uploaded a 90 minute tutorial, which is designed to be the one place I point coders at when they ask "hey, tell me everything I need to know about LLMs!" It starts at the basics: the 3-step pre-training / fine-tuning / classifier ULMFiT approach used in all modern LLMs.
Happy to share that CoDi, Any-to-Any generation, is accepted at NeurIPS! See y'all in NOLA🎷 Code & Model: github.com/microsoft/i-Co… #NeurIPS2023 @ZinengTang @mohitban47 @uncnlp
github.com
GitHub - microsoft/i-Code
Contribute to microsoft/i-Code development by creating an account on GitHub.
Any-to-Any Generation via Composable Diffusion present Composable Diffusion (CoDi), a novel generative model capable of generating any combination of output modalities, such as language, image, video, or audio, from any combination of input modalities paper page:…
I would like to inform everyone at Notre Dame, including all haters and losers (of which, sadly, there are many) that this is happening later today!
United States เทรนด์
- 1. Happy Halloween 718K posts
- 2. Dolphins 39.9K posts
- 3. YouTube TV 46.9K posts
- 4. Ryan Rollins 11.9K posts
- 5. #SinisterMinds 5,633 posts
- 6. Ravens 55.3K posts
- 7. Mike McDaniel 4,980 posts
- 8. Lamar 51.7K posts
- 9. Talbot N/A
- 10. Mary Ann N/A
- 11. Derrick Henry 5,473 posts
- 12. YTTV N/A
- 13. Hulu 18.6K posts
- 14. #DBX4 1,759 posts
- 15. #RHOC 3,322 posts
- 16. Achane 4,786 posts
- 17. UTSA 3,669 posts
- 18. #TNFonPrime 2,970 posts
- 19. Corey Perry N/A
- 20. #PorVida 1,923 posts
คุณอาจชื่นชอบ
-
Kalyan KS
@kalyan_kpl -
Nihaar Shah
@non_gaussian -
Jake Stark
@starktweetsgood -
Michael Goin
@mgoin_ -
DHANUSH ARAVINDHAN S
@Dhanush2kkid -
Aditya Nambiar
@anambiar007 -
Evil Genius
@evil_genius_v12 -
Hsuan Su
@jacksukk -
Tariq Abdalah Ali
@foster3232 -
Ayush Bihani
@hsuyab -
Yiorgos Bagakis
@yiorgosbagakis -
Renat
@Renat_Abbiazov -
Vadim G
@vadimg73 -
Carolina
@xaosxrusotos
Something went wrong.
Something went wrong.