TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Adaptive RAG – dynamic retrieval methods adjustment

126 pointsby milliondreamsabout 1 year ago

7 comments

whakimabout 1 year ago
From a consumer perspective, this is a super interesting paper because it touches on one of the fundamental issues with most RAG beyond the toy case - that you need to do different stuff depending on what the user is asking for. You also (usually) can&#x27;t just ask because most users don&#x27;t know that LLMs are bad at math or semantic search won&#x27;t be sufficient to answer questions that involve enumeration or totality. And while you can always add more steps to your RAG pipeline, some of those steps may be computationally expensive or not particularly relevant to the question at hand.<p>That being said, it is a bit frustrating that so much RAG research focuses on multi-hop approaches with LLMs. IME multiple round trips to an LLM is essentially a non-starter for any serious consumer product as it&#x27;s far too slow. Smaller models can struggle to follow instructions so they often can&#x27;t be an adequate replacement even for simpler tasks. Curious to hear if other folks working in this space have had any success thinking critically about these types of problems!
评论 #39893811 未加载
humansareok1about 1 year ago
Is there any real point to further RAG work given extremely large contexts are clearly on the way with 1M token contexts already proven?
评论 #39894630 未加载
评论 #39899132 未加载
评论 #39894530 未加载
评论 #39894625 未加载
评论 #39894644 未加载
评论 #39895385 未加载
评论 #39894674 未加载
jonnycoderabout 1 year ago
This seems similar to building a RAG router (1) to perform dynamic retrieval&#x2F;querying over data.<p>After getting hundreds of questions on my Interactive Resume AI chatbot (2), I&#x27;ve found the user queries can be categorized as: greeting, professional skills question, professional experience question, personal&#x2F;hobby question and common interview question.<p>I am currently working on building a RAG router to help improve the quality of Q&amp;A responses. I currently use gpt3.5 turbo without any special RAG techniques and the quality is lacking on performing Q&amp;A over my resume and Q&amp;A csv file. GPT4 works well but is too expensive.<p>1. <a href="https:&#x2F;&#x2F;docs.llamaindex.ai&#x2F;en&#x2F;stable&#x2F;examples&#x2F;low_level&#x2F;router&#x2F;" rel="nofollow">https:&#x2F;&#x2F;docs.llamaindex.ai&#x2F;en&#x2F;stable&#x2F;examples&#x2F;low_level&#x2F;rout...</a> 2. <a href="https:&#x2F;&#x2F;jon-olson.com&#x2F;resume_ai" rel="nofollow">https:&#x2F;&#x2F;jon-olson.com&#x2F;resume_ai</a>
machinelearningabout 1 year ago
This is a simple version of the tree search approach that people suspect Q* is
评论 #39892270 未加载
dbergabout 1 year ago
anyone know the proper github link, one in paper 404s..
评论 #39890413 未加载
oliveralbertiniabout 1 year ago
the repository links is dead
boodleboodleabout 1 year ago
Are we advertising papers on hackernews now?
评论 #39891541 未加载
评论 #39890287 未加载
评论 #39889527 未加载
评论 #39890979 未加载
评论 #39891539 未加载