<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
    <channel>
        <title><![CDATA[ Alias - pckt ]]></title>
        <link><![CDATA[ https://alias.pckt.blog ]]></link>
        <description><![CDATA[  ]]></description>
        <language>en</language>
        <pubDate>Wed, 28 Oct 2026 03:17:06 +0000</pubDate>

                    <item>
                <title>The Song of Maa</title>
                <link>https://alias.pckt.blog/the-song-of-maa-umxag9n</link>
                <description><![CDATA[edited from the recovered documents, with the anonymous Grammar (Gr) reassessed Catoid reaches us only as a written language. No speaker, no prose record, no bilingual inscription of the ordinary kind survives; what survives is one song in two documents — a text and a companion sheet — and a grammatical treatise that postdates both and depends on them. This monograph edits the song, reconstructs what can be reconstructed of the language from it, and tests the treatise claim by claim against the ...]]></description>
                <author>Alias</author>
                <guid isPermaLink="false">the-song-of-maa-umxag9n</guid>
                <pubDate>Tue, 27 Oct 2026 21:03:47 +0000</pubDate>
                            </item>
                    <item>
                <title>Classifier Evaluation</title>
                <link>https://alias.pckt.blog/classifier-evaluation-dwug23r</link>
                <description><![CDATA[Real world tasks are often unbalanced. The best baseline for comparison is nonuniform. A classifier should be evaluated on skill gain instead of raw accuracy. Looking at the distribution of classes in a benchmark dataset will reveal imbalance. We can show this using datasets lmsys/toxic-chat SetFit/sst5and ehovy/race. ToxicChat is an example of extreme imbalance. More than 90% of the examples are “benign” where the task is to identify “toxic” text.]]></description>
                <author>Alias</author>
                <guid isPermaLink="false">classifier-evaluation-dwug23r</guid>
                <pubDate>Tue, 22 Sep 2026 20:04:54 +0000</pubDate>
                            </item>
            </channel>
</rss>