<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="jiang-ferraro-2026-scribe">
    <titleInfo>
        <title>SCRIBE: Learning Reusable Skill Rubrics for Process Reward Models</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Yuxuan</namePart>
        <namePart type="family">Jiang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Francis</namePart>
        <namePart type="family">Ferraro</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>Training reliable tool-augmented agents remains challenging due to the difficulty of credit assignment in multi-step reasoning. Process Reward Models (PRMs) provide a natural solution by assigning rewards to intermediate reasoning steps or trajectories, but standard LLM-based judges often produce inconsistent signals due to underspecified evaluation criteria. Recent rubric-based reward methods improve reward quality by making scoring criteria explicit, yet they typically rely on reference answers and construct instance-level rubrics that are difficult to reuse across examples. In this work, we propose SCRIBE (Skill-Conditioned Reward with Intermediate Behavioral Evaluation), a reinforcement learning framework for reusable subgoal-level rubric rewards. SCRIBE decomposes reasoning trajectories into intermediate subgoals, clusters semantically similar reasoning behaviors into reusable Skill Prototypes, and grounds reward evaluation in prototype-level scoring criteria. This transforms reward modeling from instance-specific open-ended evaluation into reusable process-level verification. Empirical results show that SCRIBE consistently outperforms strong RL and rubric-based reward baselines across mathematical reasoning and tool-use benchmarks; notably, it improves the AIME25 score of a Qwen3-4B model from 43.3% to 63.3%. Further analysis shows that SCRIBE reduces dependence on reference answers and provides more stable, cross-instance consistent reward judgments, offering a scalable approach to reliable process reward modeling for tool-augmented agents.</abstract>
    <identifier type="citekey">jiang-ferraro-2026-scribe</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.797/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>18512</start>
            <end>18526</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T SCRIBE: Learning Reusable Skill Rubrics for Process Reward Models
%A Jiang, Yuxuan
%A Ferraro, Francis
%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 jiang-ferraro-2026-scribe
%X Training reliable tool-augmented agents remains challenging due to the difficulty of credit assignment in multi-step reasoning. Process Reward Models (PRMs) provide a natural solution by assigning rewards to intermediate reasoning steps or trajectories, but standard LLM-based judges often produce inconsistent signals due to underspecified evaluation criteria. Recent rubric-based reward methods improve reward quality by making scoring criteria explicit, yet they typically rely on reference answers and construct instance-level rubrics that are difficult to reuse across examples. In this work, we propose SCRIBE (Skill-Conditioned Reward with Intermediate Behavioral Evaluation), a reinforcement learning framework for reusable subgoal-level rubric rewards. SCRIBE decomposes reasoning trajectories into intermediate subgoals, clusters semantically similar reasoning behaviors into reusable Skill Prototypes, and grounds reward evaluation in prototype-level scoring criteria. This transforms reward modeling from instance-specific open-ended evaluation into reusable process-level verification. Empirical results show that SCRIBE consistently outperforms strong RL and rubric-based reward baselines across mathematical reasoning and tool-use benchmarks; notably, it improves the AIME25 score of a Qwen3-4B model from 43.3% to 63.3%. Further analysis shows that SCRIBE reduces dependence on reference answers and provides more stable, cross-instance consistent reward judgments, offering a scalable approach to reliable process reward modeling for tool-augmented agents.
%U https://aclanthology.org/2026.emnlp-main.797/
%P 18512-18526

Markdown (Informal)

[SCRIBE: Learning Reusable Skill Rubrics for Process Reward Models](https://aclanthology.org/2026.emnlp-main.797/) (Jiang & Ferraro, EMNLP 2026)

• SCRIBE: Learning Reusable Skill Rubrics for Process Reward Models (Jiang & Ferraro, EMNLP 2026)

ACL

• Yuxuan Jiang and Francis Ferraro. 2026. SCRIBE: Learning Reusable Skill Rubrics for Process Reward Models. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 18512–18526, Budapest, Hungary. Association for Computational Linguistics.