---
title: "Interpretable Early Termination of Web Navigation Agents via Closed Sequential Pattern Mining"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.04142v1"
permalink: "https://feed.astralor.com/items/itm_ff36562dba36af040b922c16"
published: "2026-11-03T01:35:51.000Z"
score: 57
channel: "tech"
tags: ["AI Agent","Interpretable Early Termination","Web Navigation Agents","Closed Sequential Pattern","Mining Web","Recent","Web","BIDE"]
---

# Interpretable Early Termination of Web Navigation Agents via Closed Sequential Pattern Mining

## 速览

Web navigation agents built on open-weight language models fail often and fail late, with most of the tokens spent after the trajectory has become unrecoverable. Recent work has shown that early termination is a viable …

## 关键点

- Web navigation agents built on open-weight language models fail often and fail late, with most of the tokens spent after the trajectory has become unrecoverable.
- Recent work has shown that early termination is a viable cost-reduction strategy, but existing approaches produce opaque termination decisions that do not support diagnosis or human oversight.
- In this work we apply closed sequential pattern mining (BIDE+) to K-step prefixes of agent execution traces to produce a small, human-reviewable failure-signature library that organizes into a four-category failure taxo…

## 全文

Web navigation agents built on open-weight language models fail often and fail late, with most of the tokens spent after the trajectory has become unrecoverable. Recent work has shown that early termination is a viable cost-reduction strategy, but existing approaches produce opaque termination decisions that do not support diagnosis or human oversight. In this work we apply closed sequential pattern mining (BIDE+) to K-step prefixes of agent execution traces to produce a small, human-reviewable failure-signature library that organizes into a four-category failure taxonomy with associated repair recommendations. We use a multi-variant evaluation protocol that explicitly removes contamination artifacts (zero-step crashes and trivially separable single-step successes) so that reported detection accuracy reflects behavioral discrimination rather than trace-length distributions. We evaluate our methodology on 1,544 MiniWoB++ traces collected from three open-weight language models, achieving the best F1 at K=3 on the Exclude-errors configuration across eight methods (0.717), recovering 88% of test-set failures within three steps, and supporting an early-termination policy that saves 26.3% of tokens at 92% precision.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_ff36562dba36af040b922c16