Measurement-Driven Sub-Network Selection for On-Premise Retrieval-Augmented Factory Agents
Paper recorded by Signals 4 on 2026-09-02 in cs.AI. Abstract reproduced from arXiv; link to the original below.
Published 2026-09-02 on arXiv · recorded by Signals 4 on 2026-09-03
Category: cs.AI · 人工智能 · first seen 2026-09-03
Abstract
On-premise assistants can give factory workers conversational access to machine documentation, but models capable of the task rarely fit shop-floor hardware. We show that after structural compression and retrieval-grounded adaptation, model size is no longer a reliable predictor of adapted answer quality: general capability falls almost linearly with parameter count, while judged retrieval-augment
Read on arXiv →
Cite this page: Measurement-Driven Sub-Network Selection for On-Premise Retrieval-Augmented Factory Agents: the #210 most recent of 300 cs.AI papers we have recorded (as of 2026-09-02). Source: Signals 4 (Signals API) — https://data.jiangzhang.ca/signals4/t/papers/measurement-driven-sub-network-selection-for-on-premise-retrieval-augmented-fact.html
Free to quote with attribution to “Signals 4 (Signals API)”. Machine-readable:
papers.json
Get 4 AI signals a day by email — free.
Get 4 AI signals a day by email — free