Want to make LLM-powered retrieval actually useful?
In this Agent Hour, Sarah Wooders, PhD unpacks how to build retrieval systems that deliver relevant, high-quality responses—without unnecessary complexity.
🔹 Why naive RAG often falls short and what to do instead
🔹 How fine-tuning embedding models can boost retrieval quality
🔹 When DSPi, fine-tuning, or naive RAG is the right fit for your use case
If you're working with AI retrieval (or struggling to make it work), this convo is packed with practical takeaways.
Check out the full discussion and presentation with Sarah Wooders here 👇
https://lnkd.in/gC6mUGtZ
Like, I don't think there's any model better than the LM, so why not use that to also manage memory? Yeah, OK. So that that's a fascinating idea on you're allowing the. Model to be the one that chooses what it remembers and yeah, exactly have you found that it will remember things it doesn't necessarily need to or like it misses the point sometimes or is it generally does it work well with performing those memories? I think it works pretty well. Yeah. I mean the mention T paper was written like a year ago and I I think yeah, it's kind of like stood the test of time in terms of just continuing to work really well. I think part of why it works so well is because Lemmy's kind of just speak text, right? So like it sounds like a little jank like just writing text as your memory, but that is in some ways like the most like friendly memory representation for all levels. We do have like some additional work we're doing in this space. So we are starting to we are we're kind of introducing a new way to manage. Memory where we actually have like a multi agent setup, like one agent that specializes in memory management, managing the memory of the other agent. So we're seeing better results from that over the long term because that agent will actually have the ability to kind of, you know, go back and revise its entirety of its memory as opposed to the MPT agent can only kind of make like, you know, incremental changes to its memory.
MLOps | DevOps | LLMs | LLMOps | Lecturer
17hReally appreciate how clearly Sarah Wooders, PhD broke down RAG vs. fine-tuning 🤩 so many practical takeaways here. What’s worked best for you so far?