Saved in:
Bibliographic Details
Main Authors: Cohen, Danielle, Halpern, Yoni, Kahlon, Noam, Oren, Joel, Berkovitch, Omri, Caduri, Sapir, Dagan, Ido, Efros, Anatoly
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.12423
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916950835200000
author Cohen, Danielle
Halpern, Yoni
Kahlon, Noam
Oren, Joel
Berkovitch, Omri
Caduri, Sapir
Dagan, Ido
Efros, Anatoly
author_facet Cohen, Danielle
Halpern, Yoni
Kahlon, Noam
Oren, Joel
Berkovitch, Omri
Caduri, Sapir
Dagan, Ido
Efros, Anatoly
contents Understanding user intents from UI interaction trajectories remains a challenging, yet crucial, frontier in intelligent agent development. While massive, datacenter-based, multi-modal large language models (MLLMs) possess greater capacity to handle the complexities of such sequences, smaller models which can run on-device to provide a privacy-preserving, low-cost, and low-latency user experience, struggle with accurate intent inference. We address these limitations by introducing a novel decomposed approach: first, we perform structured interaction summarization, capturing key information from each user action. Second, we perform intent extraction using a fine-tuned model operating on the aggregated summaries. This method improves intent understanding in resource-constrained models, even surpassing the base performance of large MLLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2509_12423
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Small Models, Big Results: Achieving Superior Intent Extraction through Decomposition
Cohen, Danielle
Halpern, Yoni
Kahlon, Noam
Oren, Joel
Berkovitch, Omri
Caduri, Sapir
Dagan, Ido
Efros, Anatoly
Artificial Intelligence
Computation and Language
Understanding user intents from UI interaction trajectories remains a challenging, yet crucial, frontier in intelligent agent development. While massive, datacenter-based, multi-modal large language models (MLLMs) possess greater capacity to handle the complexities of such sequences, smaller models which can run on-device to provide a privacy-preserving, low-cost, and low-latency user experience, struggle with accurate intent inference. We address these limitations by introducing a novel decomposed approach: first, we perform structured interaction summarization, capturing key information from each user action. Second, we perform intent extraction using a fine-tuned model operating on the aggregated summaries. This method improves intent understanding in resource-constrained models, even surpassing the base performance of large MLLMs.
title Small Models, Big Results: Achieving Superior Intent Extraction through Decomposition
topic Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2509.12423