Christof Schöch<p><span class="h-card" translate="no"><a href="https://fedihum.org/@cnDuKeli" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>cnDuKeli</span></a></span> <span class="h-card" translate="no"><a href="https://fedihum.org/@tcdh" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>tcdh</span></a></span> </p><p>The approach they used for <a href="https://fedihum.org/tags/classifying" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>classifying</span></a> the poems' function in the narrative were interesting, using <a href="https://fedihum.org/tags/LLMs" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLMs</span></a>. Performance is pretty bad so far. Keli took up the call for <a href="https://fedihum.org/tags/openness" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>openness</span></a> about <a href="https://fedihum.org/tags/failure" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>failure</span></a> from a session this morning and showed that the different models are bad in different ways, which allowed the team (and us) to learn something about the models. I think that's great and valuable! </p><p><a href="https://fedihum.org/tags/ErrorAnalysis" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ErrorAnalysis</span></a> <a href="https://fedihum.org/tags/ConfusionMatrix" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ConfusionMatrix</span></a> <a href="https://fedihum.org/tags/MajorityBaseline" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MajorityBaseline</span></a> <a href="https://fedihum.org/tags/DH2024" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>DH2024</span></a></p>