<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="csordas-etal-2023-approximating">
    <titleInfo>
        <title>Approximating Two-Layer Feedforward Networks for Efficient Transformers</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Róbert</namePart>
        <namePart type="family">Csordás</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Kazuki</namePart>
        <namePart type="family">Irie</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jürgen</namePart>
        <namePart type="family">Schmidhuber</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2023-12</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: EMNLP 2023</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Houda</namePart>
            <namePart type="family">Bouamor</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Juan</namePart>
            <namePart type="family">Pino</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Kalika</namePart>
            <namePart type="family">Bali</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Singapore</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>How to reduce compute and memory requirements of neural networks (NNs) without sacrificing performance? Many recent works use sparse Mixtures of Experts (MoEs) to build resource-efficient large language models (LMs). Here we introduce several novel perspectives on MoEs, presenting a general framework that *unifies* various methods to *approximate two-layer NNs* (e.g., feedforward blocks of Transformers), including product-key memories (PKMs). Leveraging insights from this framework, we propose methods to improve both MoEs and PKMs. Unlike prior work that compares MoEs with dense baselines under the *compute-equal* condition, our evaluation condition is *parameter-equal*, which is crucial to properly evaluate LMs. We show that our MoEs are competitive with the *dense* Transformer-XL on both the WikiText-103 and enwiki8 datasets at two different scales, while being much more resource efficient. This demonstrates that MoEs are relevant not only to extremely large LMs but also to any-scale resource-efficient LMs. Our code is public.</abstract>
    <identifier type="citekey">csordas-etal-2023-approximating</identifier>
    <identifier type="doi">10.18653/v1/2023.findings-emnlp.49</identifier>
    <location>
        <url>https://aclanthology.org/2023.findings-emnlp.49</url>
    </location>
    <part>
        <date>2023-12</date>
        <extent unit="page">
            <start>674</start>
            <end>692</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Approximating Two-Layer Feedforward Networks for Efficient Transformers
%A Csordás, Róbert
%A Irie, Kazuki
%A Schmidhuber, Jürgen
%Y Bouamor, Houda
%Y Pino, Juan
%Y Bali, Kalika
%S Findings of the Association for Computational Linguistics: EMNLP 2023
%D 2023
%8 December
%I Association for Computational Linguistics
%C Singapore
%F csordas-etal-2023-approximating
%X How to reduce compute and memory requirements of neural networks (NNs) without sacrificing performance? Many recent works use sparse Mixtures of Experts (MoEs) to build resource-efficient large language models (LMs). Here we introduce several novel perspectives on MoEs, presenting a general framework that *unifies* various methods to *approximate two-layer NNs* (e.g., feedforward blocks of Transformers), including product-key memories (PKMs). Leveraging insights from this framework, we propose methods to improve both MoEs and PKMs. Unlike prior work that compares MoEs with dense baselines under the *compute-equal* condition, our evaluation condition is *parameter-equal*, which is crucial to properly evaluate LMs. We show that our MoEs are competitive with the *dense* Transformer-XL on both the WikiText-103 and enwiki8 datasets at two different scales, while being much more resource efficient. This demonstrates that MoEs are relevant not only to extremely large LMs but also to any-scale resource-efficient LMs. Our code is public.
%R 10.18653/v1/2023.findings-emnlp.49
%U https://aclanthology.org/2023.findings-emnlp.49
%U https://doi.org/10.18653/v1/2023.findings-emnlp.49
%P 674-692

Markdown (Informal)

[Approximating Two-Layer Feedforward Networks for Efficient Transformers](https://aclanthology.org/2023.findings-emnlp.49) (Csordás et al., Findings 2023)

• Approximating Two-Layer Feedforward Networks for Efficient Transformers (Csordás et al., Findings 2023)

ACL

• Róbert Csordás, Kazuki Irie, and Jürgen Schmidhuber. 2023. Approximating Two-Layer Feedforward Networks for Efficient Transformers. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 674–692, Singapore. Association for Computational Linguistics.