The scientist who coined retrieval augmented generation

Dr. Patrick Lewis, who coined the term RAG (Retrieval Augmented Generation) and now works at Cohere, discusses the evolution of language models, RAG systems, and challenges in AI evaluation.

MLST is sponsored by Brave:

Key topics covered:
- Origins and evolution of Retrieval Augmented Generation (RAG)
- Challenges in evaluating RAG systems and language models
- Human-AI collaboration in research and knowledge work
- Word embeddings and the progression to modern language models
- Dense vs sparse retrieval methods in information retrieval

The discussion also explored broader implications and applications:
- Balancing faithfulness and fluency in RAG systems
- User interface design for AI-augmented research tools
- The journey from chemistry to AI research
- Challenges in enterprise search compared to web search
- The importance of data quality in training AI models

Cohere Command Models, check them out - they are amazing for RAG!

TOC
00:00:00 1. Intro to RAG
00:05:30 2. RAG Evaluation: Poll framework & model performance
00:12:55 3. Data Quality: Cleanliness vs scale in AI training
00:15:13 4. Human-AI Collaboration: Research agents & UI design
00:22:57 5. RAG Origins: Open-domain QA to generative models
00:30:18 6. RAG Challenges: Info retrieval, tool use, faithfulness
00:42:01 7. Dense vs Sparse Retrieval: Techniques & trade-offs
00:47:02 8. RAG Applications: Grounding, attribution, hallucination prevention
00:54:04 9. UI for RAG: Human-computer interaction & model optimization
00:59:01 10. Word Embeddings: Word2Vec, GloVe, and semantic spaces
01:06:43 11. Language Model Evolution: BERT, GPT, and beyond
01:11:38 12. AI & Human Cognition: Sequential processing & chain-of-thought

Refs:
1. Retrieval Augmented Generation (RAG) paper / Patrick Lewis et al. [00:27:45]
2. LAMA (LAnguage Model Analysis) probe / Petroni et al. [00:26:35]
3. KILT (Knowledge Intensive Language Tasks) benchmark / Petroni et al. [00:27:05]
4. Word2Vec algorithm / Tomas Mikolov et al. [01:00:25]
5. GloVe (Global Vectors for Word Representation) / Pennington et al. [01:04:35]
6. BERT (Bidirectional Encoder Representations from Transformers) / Devlin et al. [01:08:00]
7. 'The Language Game' book / Nick Chater and Morten H. Christiansen [01:11:40]

Disclaimer: This is the sixth video from our Cohere partnership. We were not told what to say in the interview. Filmed in London in June 2024.

[sonOfLiberty]

Wow Machine Learning become such an awesome podcast. It was it before, but you really leveled up

[psij]

You guys have taken this to the next level 🎉

[diga]

I also see vision and other temporaly adjacent modalities used in training and fine-tuning being very useful in reasoning depth.

[ronaldevans]

I think this is much bigger than most people in the AI space realize.

[nonchai]

I get the impression that Copilot these days - in the browser- is also doing RAG? - has copilot ( aided byBing ) upped its game in recent months? It seems nearer to Perplexity now than being just a LLM

[egor.okhterov]

Does feeling something require consciousness?
Does feeling something induce consciousness?
Can consciousness exist without feeling?
What is the interplay between feelings and consciousness?

[jameswhitaker]

Always love these.! My favorite pod. Bro needs a hair change tho

[goldnutter]

Retrieval rules for safety buff ?

BTW Gonna need an oracle for that.. look who is an advisor to Chainlink.. Eric Schmidt

Amazing years ahead we just need to be responsible and keep the "it's not about me" ethos

[_ARCATEC_]

SpaceTime Vector Continuum. •X(s zVc qt() ZC ()TQ zvc S)Y•