🚀🔍 AI papers deep dive: LLM understanding, RAG, CoT

🌐 RA-DIT Retrieval-Augmented Dual Instruction Tuning: Enhancing RAG systems by fine-tuning both the LM & retriever. Implementation in LlamaIndex.
📆 FreshLLM: Leveraging search engines to keep LLMs updated with the latest info.
🧠 Analogical Reasoners: Automatically guides the reasoning of LLMs for a better chain-of-thought.
✨ On AI transparency & interpretability, two pivotal papers:
- Representation Engineering: A Top-Down Approach to AI Transparency
- Decomposing Language Models Into Understandable Components
#ArtificialIntelligence #AIpapers
Links:
arxiv.org/abs/2310.01352
/ 1709646787076935818
arxiv.org/pdf/2310.03214v1.pdf
arxiv.org/pdf/2310.01714.pdf
www.ai-transparency.org/
www.anthropic.com/index/decom...
transformer-circuits.pub/2023...
00:00 intro
01:04 RA-DIT Retrieval-Augmented Dual Instruction Tuning
04:17 FreshLLMs
07:02 Analogical Reasoners
10:07 Representation Engineering
14:18 Decomposing Language Models Into Understandable Components
🔔 SUBSCRIBE to my channel: kzbin.info...
⭐ Stay in touch ⭐
📚 DS/ML Book Club: dsbookclub.github.io/
▶ KZbin: / sophiayangds
✍️ Medium: / sophiamyang
🐦 Twitter: / sophiamyang
🤝 Linkedin: / sophiamyang
💚 #ai

[@VictoriaLin]

We're flattered to be the first paper featured in your first paper reading session. Continue the great work! 👩🏻‍💻📖✨

[@SophiaYangDS]

Thanks so much! It's an honor to get a comment from the author yourself! ♥

[@sgramstrup]

I think its a good idea. We all need some 'bridging' between the most important new papers, and all of us that are a bit swamped by the speed of this field. Also, by discussing the details of the paper, the audience gets a referential framework to relate concepts to. Keep digging for papers that reveal the inner workings of 'cognitive networks'..

[@alexeymalafeev6167]

Awesome video!

[@santhoshsthanikam9656]

This is excellent, Please continue to make this type of videos.Thanks for sharing the knowledge

[@SophiaYangDS]

Thank you 🙏 I just made a new one. Hope you like it :)

[@snsa_kscc]

Great stuff, Sophia! Have a good one.

[@SophiaYangDS]

Thanks!

[@FREELEARNING]

Thanks Sophia. What I like the most about your videos is that they are concise.

[@SophiaYangDS]

Thanks so much 🙏

[@kostastsaousis7445]

Sophia. This session is amazing and a very good idea! Please continue to do it as often as possible!! Great work !

[@ark1111]

Great Video and very intresting papers you found! Thank you for showcasing them!

[@SophiaYangDS]

Thank you 🙏

[@dimastorres8530]

Representation engineering was a mind blower. Thank you for sharing!

[@SophiaYangDS]

Thank you 🙏

[@AIJasonZ]

Thank you Sophia, so good! How did you go and find quality paper to read? (I mean there are so many new ones each week LOL)

[@SophiaYangDS]

I follow a bunch of researchers and people who share quality papers 😄

[@oscarbertel1449]

Excelente

[@SophiaYangDS]

Thank you 🙏

[@VenkatesanVenkat-fd4hg]

Superr, explanation

[@SophiaYangDS]

Thank you!

[@densonsmith2]

Very useful

[@SophiaYangDS]

Thank you 🙏

[@klammer75]

Great work and love the paper reviews! Keep em coming!🥳😎🦾

[@viewview6687]

Thank you very much for this video. I watch your summarize video before bed.

[@SophiaYangDS]

Thanks so much! This one is a little different from my other videos. I hope it's not too boring 😅

[@leo-ur8kx]

@@SophiaYangDS Its not boring at all. Do you planning to create video that shows reasoning in embedding search?

[@SophiaYangDS]

Thank you!!! Really good to know. And thanks for the suggestions. Do you have recommended paper on that? @@leo-ur8kx

[@VaibhavPatil-rx7pc]

Thanks

[@SophiaYangDS]

Thanks!

[@moon8013]

keep making videos please

[@SophiaYangDS]

Thanks 🙏 I'll try