Smart LLM Routing & Novel Multi-Agent Framework

New AI system will recommend the perfect AI model for your specific task at your specific price point. No more try and error? The perfect AI? Smile.

And why should we code at all? New AI systems code End-to-End for AI apps and decide autonomously on what LLMs to choose, how to fine-tune and when to optimize the overall planning, given a multi-agent system and intense optimization procedures. All based on artificial intelligence of multiple LLMs and VLMs.

Should we trust those end-to-end AI systems? And would any recommendation by those AI system be biased? Maybe towards its creator /global corporation?

All rights with Authors:
Graph Router : A GRAPH-BASED ROUTER FOR LLM SELECTIONS

AUTOML-AGENT: A MULTI-AGENT LLM
FRAMEWORK FOR FULL-PIPELINE AUTOML

00:00 Smart LLM Routing
04:14 How to build a GRAPH representation
09:49 From a Graph to a Graph NN
14:04 Multi-Agent LLM for full pipeline
18:53 Retrieval and Coding pushed to the side
25:41 New Focus for End-to-End AI

#airesearch
#aiagents
#coding

[dennis]

From a real world product perspective a graph ML approach towards effective prompt routing is completely over engineered. The underlying problem is simple and given the speed and intensity of ongoing progress a solution must consider performance and adaptability at the same time.

[johnkintree]

GraphRouter is an interesting concept. A Wiki implementation that would allow many people to input the performance of different LLMs on different tasks might be helpful. Users should also be able to specify their hardware, such as Oneplus 11 smartphone with 16 GB RAM, so the GraphRouter would select an LLM that is optimized for it.

[dragoon]

Its not this hard...

Distributed neural network. You add on tensors onto each edge device depending how much they can handle then you distribute the computing across many many many devices. You can scale this infinitely. Its like a bonnet neural net.

[Philosophizer]

A neural network trained to find novel neural networks .