@inproceedings{simhadri-2026-capability,
title = "Capability Is Not Complementarity: A Routing Feasibility Study in Financial Question Answering",
author = "Simhadri, Rishi N.",
editor = "Kamalloo, Ehsan and
Gontier, Nicolas and
Rozanov, Nikolai and
Hashemzadeh, Maryam and
Arora, Akhil",
booktitle = "Proceedings of the Second Workshop for {RE}search on Agent Language Models ({REALM} 2026)",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.realm-1.37/",
pages = "551--556",
ISBN = "979-8-89176-485-9",
abstract = "Selective routing is useful only when an expensive branch solves cases that a cheaper branch misses. We call this prerequisite expert complementarity and ask whether a capable iterative agent actually provides it. In a prospectively frozen financial question answering study, we compare Static with bounded ReAct across multiple capability tiers. Within each tier, the arms share the model, answer contract, retrieval interface, scoring, and resource limits. A stronger model removes ReAct{'}s average deficit and brings it to parity with Static, showing that relative performance depends on model choice. Yet Static and ReAct still succeed on nearly the same items. Across 150 shared items per tier, ReAct adds only one or two unique exact successes; oracle quality headroom point estimates are at most 1.7 points, while ReAct takes 2.1 to 2.6 times as long. A stability check preserves this pattern. Capability and complementarity are therefore distinct: improving an expensive branch can make it competitive without making it useful for routing. Evaluation contracts should be harmonized and paired complementarity audited before training a selector. This result concerns bounded ReAct over controlled benchmark evidence, not iterative agents in general."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="simhadri-2026-capability">
<titleInfo>
<title>Capability Is Not Complementarity: A Routing Feasibility Study in Financial Question Answering</title>
</titleInfo>
<name type="personal">
<namePart type="given">Rishi</namePart>
<namePart type="given">N</namePart>
<namePart type="family">Simhadri</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 Second Workshop for REsearch on Agent Language Models (REALM 2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ehsan</namePart>
<namePart type="family">Kamalloo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nicolas</namePart>
<namePart type="family">Gontier</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nikolai</namePart>
<namePart type="family">Rozanov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Maryam</namePart>
<namePart type="family">Hashemzadeh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Akhil</namePart>
<namePart type="family">Arora</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">979-8-89176-485-9</identifier>
</relatedItem>
<abstract>Selective routing is useful only when an expensive branch solves cases that a cheaper branch misses. We call this prerequisite expert complementarity and ask whether a capable iterative agent actually provides it. In a prospectively frozen financial question answering study, we compare Static with bounded ReAct across multiple capability tiers. Within each tier, the arms share the model, answer contract, retrieval interface, scoring, and resource limits. A stronger model removes ReAct’s average deficit and brings it to parity with Static, showing that relative performance depends on model choice. Yet Static and ReAct still succeed on nearly the same items. Across 150 shared items per tier, ReAct adds only one or two unique exact successes; oracle quality headroom point estimates are at most 1.7 points, while ReAct takes 2.1 to 2.6 times as long. A stability check preserves this pattern. Capability and complementarity are therefore distinct: improving an expensive branch can make it competitive without making it useful for routing. Evaluation contracts should be harmonized and paired complementarity audited before training a selector. This result concerns bounded ReAct over controlled benchmark evidence, not iterative agents in general.</abstract>
<identifier type="citekey">simhadri-2026-capability</identifier>
<location>
<url>https://aclanthology.org/2026.realm-1.37/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>551</start>
<end>556</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Capability Is Not Complementarity: A Routing Feasibility Study in Financial Question Answering
%A Simhadri, Rishi N.
%Y Kamalloo, Ehsan
%Y Gontier, Nicolas
%Y Rozanov, Nikolai
%Y Hashemzadeh, Maryam
%Y Arora, Akhil
%S Proceedings of the Second Workshop for REsearch on Agent Language Models (REALM 2026)
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-485-9
%F simhadri-2026-capability
%X Selective routing is useful only when an expensive branch solves cases that a cheaper branch misses. We call this prerequisite expert complementarity and ask whether a capable iterative agent actually provides it. In a prospectively frozen financial question answering study, we compare Static with bounded ReAct across multiple capability tiers. Within each tier, the arms share the model, answer contract, retrieval interface, scoring, and resource limits. A stronger model removes ReAct’s average deficit and brings it to parity with Static, showing that relative performance depends on model choice. Yet Static and ReAct still succeed on nearly the same items. Across 150 shared items per tier, ReAct adds only one or two unique exact successes; oracle quality headroom point estimates are at most 1.7 points, while ReAct takes 2.1 to 2.6 times as long. A stability check preserves this pattern. Capability and complementarity are therefore distinct: improving an expensive branch can make it competitive without making it useful for routing. Evaluation contracts should be harmonized and paired complementarity audited before training a selector. This result concerns bounded ReAct over controlled benchmark evidence, not iterative agents in general.
%U https://aclanthology.org/2026.realm-1.37/
%P 551-556
Markdown (Informal)
[Capability Is Not Complementarity: A Routing Feasibility Study in Financial Question Answering](https://aclanthology.org/2026.realm-1.37/) (Simhadri, REALM 2026)
ACL