@inproceedings{li-li-2026-browsecomp,
title = "{B}rowse{C}omp-Verify: Benchmarking Self-derived Verification Agents in Web Search",
author = "Li, Guanzhen and
Li, Shijie",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.1889/",
pages = "43046--43075",
ISBN = "979-8-89176-459-0",
abstract = "Verification agents are increasingly used to judge candidate answers in web search, yet their own reliability remains underexplored. Existing verification methods often rely on annotated ground truths or manually crafted rubrics, limiting scalability under high annotation costs and the dynamic nature of the web. Hence, we propose the self-derived verification task, where an agent derives validation rules from a search query and verifies whether a candidate answer satisfies them using external tools, without human-provided rubrics or ground-truth answers. We further introduce BrowseComp-Verify, a benchmark of 3298 query-answer pairs constructed from 453 BrowseComp questions with reliable correctness labels. Using BrowseComp-Verify, we evaluate a broad suite of LLMs as self-derived verification agents. Most models complete the task with high coverage, but accurate judgment remains challenging. In particular, many models exhibit a conservative bias toward rejecting candidate answers: GPT-v5.2-Thinking-high achieves 99.04{\%} accuracy on incorrect answers but only 20.39{\%} on correct answers."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-li-2026-browsecomp">
<titleInfo>
<title>BrowseComp-Verify: Benchmarking Self-derived Verification Agents in Web Search</title>
</titleInfo>
<name type="personal">
<namePart type="given">Guanzhen</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shijie</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Verification agents are increasingly used to judge candidate answers in web search, yet their own reliability remains underexplored. Existing verification methods often rely on annotated ground truths or manually crafted rubrics, limiting scalability under high annotation costs and the dynamic nature of the web. Hence, we propose the self-derived verification task, where an agent derives validation rules from a search query and verifies whether a candidate answer satisfies them using external tools, without human-provided rubrics or ground-truth answers. We further introduce BrowseComp-Verify, a benchmark of 3298 query-answer pairs constructed from 453 BrowseComp questions with reliable correctness labels. Using BrowseComp-Verify, we evaluate a broad suite of LLMs as self-derived verification agents. Most models complete the task with high coverage, but accurate judgment remains challenging. In particular, many models exhibit a conservative bias toward rejecting candidate answers: GPT-v5.2-Thinking-high achieves 99.04% accuracy on incorrect answers but only 20.39% on correct answers.</abstract>
<identifier type="citekey">li-li-2026-browsecomp</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1889/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>43046</start>
<end>43075</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T BrowseComp-Verify: Benchmarking Self-derived Verification Agents in Web Search
%A Li, Guanzhen
%A Li, Shijie
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F li-li-2026-browsecomp
%X Verification agents are increasingly used to judge candidate answers in web search, yet their own reliability remains underexplored. Existing verification methods often rely on annotated ground truths or manually crafted rubrics, limiting scalability under high annotation costs and the dynamic nature of the web. Hence, we propose the self-derived verification task, where an agent derives validation rules from a search query and verifies whether a candidate answer satisfies them using external tools, without human-provided rubrics or ground-truth answers. We further introduce BrowseComp-Verify, a benchmark of 3298 query-answer pairs constructed from 453 BrowseComp questions with reliable correctness labels. Using BrowseComp-Verify, we evaluate a broad suite of LLMs as self-derived verification agents. Most models complete the task with high coverage, but accurate judgment remains challenging. In particular, many models exhibit a conservative bias toward rejecting candidate answers: GPT-v5.2-Thinking-high achieves 99.04% accuracy on incorrect answers but only 20.39% on correct answers.
%U https://aclanthology.org/2026.emnlp-main.1889/
%P 43046-43075
Markdown (Informal)
[BrowseComp-Verify: Benchmarking Self-derived Verification Agents in Web Search](https://aclanthology.org/2026.emnlp-main.1889/) (Li & Li, EMNLP 2026)
ACL