Anish Pathak

[Anish Pathak @AnishPathak2206]

Model: https://t.co/HZxenNnb4h
Code: https://t.co/fElMpl6h7b
Paper: coming soon.
Feedback welcome, especially where the date rules or the fixed function list break on your data.

[Anish Pathak @AnishPathak2206]

A model 60 times smaller beat Google's FunctionGemma on Mobile Actions.

ARN: 4.4M parameters vs 268M. Same training data. 11 ms on one CPU thread.

[Anish Pathak @AnishPathak2206]

The caveats, up front:

I looked at the test set while developing, so these results are exploratory.
With no rules, on a separate test, ARN reaches 40.9 on MixATIS and 70.3 on MixSNIPS. That is below several published models.
The full limitations are in the repo.

[Anish Pathak @AnishPathak2206]

I just published Building a Persistent Codebase Memory System: Architecture, Hybrid Graph RAG, and Lessons Learned https://t.co/DRCdN7qTYv

[Anish Pathak @AnishPathak2206]

I just published How I Built AETHER: A Local AI Assistant That Controls My PC, Sends WhatsApp Messages, and Learns… https://t.co/6J9hNTDEcy
Thanks to @towards_AI for reviewing and accepting my article.