SKILL: Structured Knowledge Infusion for Large Language Models · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2205.08184/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTSKILL: Structured Knowledge Infusion for Large Language Models — mentioned by webmavenhttps://news.ycombinator.com/item?id=34591408urn:hackerpapers:mention:2205.08184:34591408Tue, 31 Jan 2023 05:05:52 GMT<p>&gt; There are plenty of other things I&#x27;ve searched for recently, were ChatGPT simply doesn&#x27;t make sense. One example is looking for trusted reviews of kitchen appliances, and things to do on vacation. Right now, ChatGPT doesn&#x27;t provide good answers to questions in the form of &quot;What is a good source for information of type X?&quot;, but that is really a limitation of the source data (text from a current web crawl will, superficially, tend to reflect information rather than meta analysis), rather than the model as such, and transforming PageRank style graph data into a verbalized form suitable for an…</p><p><a href="https://hackerpaper.kaipfister.de/p/2205.08184/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.CL