<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="swaminathan-etal-2026-scitrace">
    <titleInfo>
        <title>SciTrace: Trajectory-Aware Safety Reasoning for Scientific Discovery Agents</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Tanush</namePart>
        <namePart type="family">Swaminathan</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Runmin</namePart>
        <namePart type="family">Jiang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Letian</namePart>
        <namePart type="family">Zhang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Min</namePart>
        <namePart type="family">Xu</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>Findings of the Association for Computational Linguistics: EMNLP 2026</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">979-8-89176-462-0</identifier>
    </relatedItem>
    <abstract>LLM-based scientific agents have shown strong capacity for autonomous research, yet their safety layers remain structurally divorced from core reasoning: they inspect pipeline outputs rather than shaping the deliberation that produces them. This separation opens two failure modes: safety signals accumulated at one stage are discarded before the next, and sequences of individually benign tool calls can compose into harmful outcomes that no single-step filter detects. To address these challenges, we introduce **SciTrace**, a framework that weaves safety reasoning into every stage of the scientific agent pipeline. SciTrace couples two complementary mechanisms: a *Safety-Intrinsic Reasoning Loop* (SIR) that maintains a cumulative risk state across the Thinker, Experimenter, Writer, and Reviewer stages through joint task-and-safety deliberation, and a *Compositional Tool-Chain Verifier* (CTV) that performs trajectory-aware safety checks before execution, catching risks that surface only across multi-step tool sequences. Evaluated on 240 high-risk research tasks and 120 tool-related risk tasks spanning six scientific domains, SciTrace achieves state-of-the-art (**SOTA**) safety among compared frameworks across four backbone models: it consistently improves tool call safety and adversarial robustness while preserving scientific output quality, and it uncovers **78.8%** of the compositional tool-chain escapes that single-step monitors miss. The project website is available at [https://opensciagent.github.io/SciTrace/](https://opensciagent.github.io/SciTrace/).</abstract>
    <identifier type="citekey">swaminathan-etal-2026-scitrace</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-emnlp.1502/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>31988</start>
            <end>32010</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T SciTrace: Trajectory-Aware Safety Reasoning for Scientific Discovery Agents
%A Swaminathan, Tanush
%A Jiang, Runmin
%A Zhang, Letian
%A Xu, Min
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F swaminathan-etal-2026-scitrace
%X LLM-based scientific agents have shown strong capacity for autonomous research, yet their safety layers remain structurally divorced from core reasoning: they inspect pipeline outputs rather than shaping the deliberation that produces them. This separation opens two failure modes: safety signals accumulated at one stage are discarded before the next, and sequences of individually benign tool calls can compose into harmful outcomes that no single-step filter detects. To address these challenges, we introduce **SciTrace**, a framework that weaves safety reasoning into every stage of the scientific agent pipeline. SciTrace couples two complementary mechanisms: a *Safety-Intrinsic Reasoning Loop* (SIR) that maintains a cumulative risk state across the Thinker, Experimenter, Writer, and Reviewer stages through joint task-and-safety deliberation, and a *Compositional Tool-Chain Verifier* (CTV) that performs trajectory-aware safety checks before execution, catching risks that surface only across multi-step tool sequences. Evaluated on 240 high-risk research tasks and 120 tool-related risk tasks spanning six scientific domains, SciTrace achieves state-of-the-art (**SOTA**) safety among compared frameworks across four backbone models: it consistently improves tool call safety and adversarial robustness while preserving scientific output quality, and it uncovers **78.8%** of the compositional tool-chain escapes that single-step monitors miss. The project website is available at [https://opensciagent.github.io/SciTrace/](https://opensciagent.github.io/SciTrace/).
%U https://aclanthology.org/2026.findings-emnlp.1502/
%P 31988-32010

Markdown (Informal)

[SciTrace: Trajectory-Aware Safety Reasoning for Scientific Discovery Agents](https://aclanthology.org/2026.findings-emnlp.1502/) (Swaminathan et al., Findings 2026)

• SciTrace: Trajectory-Aware Safety Reasoning for Scientific Discovery Agents (Swaminathan et al., Findings 2026)

ACL

• Tanush Swaminathan, Runmin Jiang, Letian Zhang, and Min Xu. 2026. SciTrace: Trajectory-Aware Safety Reasoning for Scientific Discovery Agents. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 31988–32010, Budapest, Hungary. Association for Computational Linguistics.