Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
sigmoid.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A social space for people researching, working with, or just interested in AI!

Administered by:

Server stats:

649
active users

sigmoid.social: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.8

#eacl2023

0 posts0 participants0 posts today
Tom Stafford<p>Chatbots for Good and Evil</p><p><a href="https://underline.io/lecture/72154-chatbots-for-good-and-evil" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">underline.io/lecture/72154-cha</span><span class="invisible">tbots-for-good-and-evil</span></a> </p><p>Kevin Munger keynote at <a href="https://mastodon.online/tags/EACL2023" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>EACL2023</span></a></p>
Venkat<p>I’m at <a href="https://sigmoid.social/tags/eacl2023" class="mention hashtag" rel="tag">#<span>eacl2023</span></a> presenting my work on intergroup bias :) Come by and say hi <a href="https://underline.io/events/383/posters/14966/poster/71341-how-people-talk-about-each-other-modeling-generalized-intergroup-bias-and-emotion" target="_blank" rel="nofollow noopener noreferrer" translate="no"><span class="invisible">https://</span><span class="ellipsis">underline.io/events/383/poster</span><span class="invisible">s/14966/poster/71341-how-people-talk-about-each-other-modeling-generalized-intergroup-bias-and-emotion</span></a><br /><a href="https://aclanthology.org/2023.eacl-main.183/" target="_blank" rel="nofollow noopener noreferrer" translate="no"><span class="invisible">https://</span><span class="ellipsis">aclanthology.org/2023.eacl-mai</span><span class="invisible">n.183/</span></a></p>
marco<p>I&#39;ll be attending (and presenting a poster at) <a href="https://sigmoid.social/tags/EACL" class="mention hashtag" rel="tag">#<span>EACL</span></a> this week. Definitely reach out if you want to grab coffee!</p><p>My poster is</p><p>`Parameter-Efficient Korean Character-Level Language Modeling`,</p><p>where we describe a way to efficiently encode and decode (Korean) syllable-level representations without requiring an embedding for each possible syllable in the Korean writing system (more than 11k).</p><p><a href="https://sigmoid.social/tags/nlp" class="mention hashtag" rel="tag">#<span>nlp</span></a> <a href="https://sigmoid.social/tags/nlproc" class="mention hashtag" rel="tag">#<span>nlproc</span></a> <a href="https://sigmoid.social/tags/eacl2023" class="mention hashtag" rel="tag">#<span>eacl2023</span></a> <a href="https://sigmoid.social/tags/cjk" class="mention hashtag" rel="tag">#<span>cjk</span></a></p>
Venkat<p>Super happy to announce that my paper studying interpersonal bias with <span class="h-card" translate="no"><a href="https://sigmoid.social/@jessyjli" class="u-url mention">@<span>jessyjli</span></a></span> , David Beaver, Malihe Alikhani, Katherine Atwell and Barea Sinno has been accepted to <a href="https://sigmoid.social/tags/eacl2023" class="mention hashtag" rel="tag">#<span>eacl2023</span></a>.I’m really excited for this first step in my research thesis, and for what I&#39;ll find in the months to come.<br />Our goal is to study how the relationship between the speaker and the target of an utterance influences the language itself. Language is biased regardless of intent , and studying the effect of interpersonal…🧵1/3</p>
eaclmeeting<p>Official dates for <a href="https://sigmoid.social/tags/eacl2023" class="mention hashtag" rel="tag">#<span>eacl2023</span></a> !</p><p>May 2, 3, 4 - Main Conference<br />May 5 &amp; 6 - Workshops and Tutorials </p><p>Looking forward to seeing everyone in three months!</p><p><a href="https://sigmoid.social/tags/NLProc" class="mention hashtag" rel="tag">#<span>NLProc</span></a></p>
Hosein Mohebbi<p>🥳 Thrilled to announce our paper got accepted to <a href="https://sigmoid.social/tags/EACL2023" class="mention hashtag" rel="tag">#<span>EACL2023</span></a>!<br />We introduce *Value Zeroing*, a new interpretability method for quantifying context mixing in Transformers.</p><p>A joint work w/ me, <span class="h-card" translate="no"><a href="https://sigmoid.social/@wzuidema" class="u-url mention">@<span>wzuidema</span></a></span>, <span class="h-card" translate="no"><a href="https://sigmoid.social/@gchrupala" class="u-url mention">@<span>gchrupala</span></a></span>, and Afra</p><p>📑Paper: <a href="https://arxiv.org/abs/2301.12971" target="_blank" rel="nofollow noopener noreferrer" translate="no"><span class="invisible">https://</span><span class="">arxiv.org/abs/2301.12971</span><span class="invisible"></span></a><br />☕Code: <a href="https://github.com/hmohebbi/ValueZeroing" target="_blank" rel="nofollow noopener noreferrer" translate="no"><span class="invisible">https://</span><span class="ellipsis">github.com/hmohebbi/ValueZeroi</span><span class="invisible">ng</span></a></p><p><a href="https://sigmoid.social/tags/NLProc" class="mention hashtag" rel="tag">#<span>NLProc</span></a> <a href="https://sigmoid.social/tags/InDeep" class="mention hashtag" rel="tag">#<span>InDeep</span></a></p>
UKP Lab<p>Six papers authored or co-authored by UKP members have been accepted for publication at <a href="https://sigmoid.social/tags/EACL2023" class="mention hashtag" rel="tag">#<span>EACL2023</span></a>, the Conference of the European Chapter of the Association for Computational Linguistics.</p><p>The conference will be held online and in Dubrovnik, Croatia, from 2 to 6 of May 2023.</p><p>Congratulations to all involved! 💐 </p><p><a href="https://www.informatik.tu-darmstadt.de/ukp/ukp_home/news_ukp/ukp_newsdetails_273984.en.jsp" target="_blank" rel="nofollow noopener noreferrer" translate="no"><span class="invisible">https://www.</span><span class="ellipsis">informatik.tu-darmstadt.de/ukp</span><span class="invisible">/ukp_home/news_ukp/ukp_newsdetails_273984.en.jsp</span></a></p>
Zdeněk Kasner<p>Our paper with <span class="h-card" translate="no"><a href="https://sigmoid.social/@sinantie" class="u-url mention">@<span>sinantie</span></a></span> and <span class="h-card" translate="no"><a href="https://sigmoid.social/@tuetschek" class="u-url mention">@<span>tuetschek</span></a></span> got accepted to <a href="https://sigmoid.social/tags/eacl2023" class="mention hashtag" rel="tag">#<span>eacl2023</span></a>!</p><p>TL;DR: If you want pretrained models to describe data from a novel domain, use 𝘂𝗻𝗮𝗺𝗯𝗶𝗴𝘂𝗼𝘂𝘀 𝗵𝘂𝗺𝗮𝗻-𝗿𝗲𝗮𝗱𝗮𝗯𝗹𝗲 𝗹𝗮𝗯𝗲𝗹𝘀💡</p><p><a href="https://arxiv.org/abs/2210.07373" target="_blank" rel="nofollow noopener noreferrer" translate="no"><span class="invisible">https://</span><span class="">arxiv.org/abs/2210.07373</span><span class="invisible"></span></a></p><p><span class="h-card" translate="no"><a href="https://birdsite.wilde.cloud/users/ufal_cuni" class="u-url mention">@<span>ufal_cuni</span></a></span></p>
marco<p>Happy to say my first paper since moving to Tokyo to start my PhD was accepted at <a href="https://sigmoid.social/tags/EACL2023" class="mention hashtag" rel="tag">#<span>EACL2023</span></a></p><p>`Parameter-Efficient Korean Character-Level Language Modeling`</p><p>This was joint work with Sangwhan Moon (<span class="h-card" translate="no"><a href="https://hachyderm.io/@s" class="u-url mention">@<span>s</span></a></span>), Lawrence Wolf-Sonkin, and my advisor, Naoaki Okazaki.</p><p>In it, we describe a method to factor the character ( = Hangul syllables) embedding and output layers in a generic neural LM to reduce the embedding/output parameter counts by up to &gt;99% with no loss in modeling quality!</p>
NEJLT<p>NEJLT welcomes NLP papers, and offers accelerated review if you include your <a href="https://sigmoid.social/tags/EACL2023" class="mention hashtag" rel="tag">#<span>EACL2023</span></a> or <a href="https://sigmoid.social/tags/ICLR2023" class="mention hashtag" rel="tag">#<span>ICLR2023</span></a> reviews.</p><p>➡️ indexing in the ACL Anthology<br />➡️ journal-format articles<br />➡️ no length limit<br />➡️ fast, open, free review &amp; free publishing</p><p>Read more here: <a href="https://www.nejlt.org/" target="_blank" rel="nofollow noopener noreferrer" translate="no"><span class="invisible">https://www.</span><span class="">nejlt.org/</span><span class="invisible"></span></a></p>
Ivan Habernal<p>Dear <a href="https://sigmoid.social/tags/eacl2023" class="mention hashtag" rel="tag">#<span>eacl2023</span></a> reviewers - I&#39;m one of you, but I hate when you **promise to deliver** reviews after being asked several times and then nothing happens. Either say &quot;no, I&#39;m sorry&quot; or do the job. Is it that complicated?</p>
ExploreLive feeds
About