<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Ben Segal on Medium]]></title>
        <description><![CDATA[Stories by Ben Segal on Medium]]></description>
        <link>https://medium.com/@segalnyc?source=rss-74b4a6d225d3------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*9nF8fX__yNUELu-m.jpeg</url>
            <title>Stories by Ben Segal on Medium</title>
            <link>https://medium.com/@segalnyc?source=rss-74b4a6d225d3------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 11:49:26 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@segalnyc/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Structured data is foundational for the effective functioning of large language models (LLMs) and…]]></title>
            <link>https://medium.com/@segalnyc/structured-data-is-foundational-for-the-effective-functioning-of-large-language-models-llms-and-42557c401793?source=rss-74b4a6d225d3------2</link>
            <guid isPermaLink="false">https://medium.com/p/42557c401793</guid>
            <dc:creator><![CDATA[Ben Segal]]></dc:creator>
            <pubDate>Tue, 30 Apr 2024 17:07:01 GMT</pubDate>
            <atom:updated>2024-04-30T17:08:07.029Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Structured data</strong> is foundational for the effective functioning of large language models (LLMs) and generative AI systems for several reasons:</p><ol><li>Training Data: Structured data often serves as the backbone for training LLMs. While these models excel at understanding and generating natural language, they typically rely on vast amounts of structured data to learn patterns, relationships, and context. Structured data provides clear signals and examples for the model to learn from.</li><li>Feature Extraction: Structured data can be used to extract features that are relevant to the task at hand. These features can then be combined with natural language inputs to enhance the model’s understanding and generation capabilities. For example, in a language model tasked with generating product descriptions, structured data about product attributes (such as price, size, color, etc.) can be integrated to ensure coherence and accuracy.</li><li>Contextual Understanding: Structured data provides context that helps LLMs better understand the content they are generating. By incorporating information from structured data sources, the model can produce more relevant and coherent outputs. For instance, if a model is generating responses to customer inquiries, incorporating structured data about previous interactions or customer profiles can improve the relevance and helpfulness of the responses.</li><li>Data Augmentation: Structured data can be used to augment training data for LLMs, especially in cases where natural language data is limited. By synthesizing additional training examples based on structured data, models can be trained more effectively and generalize better to new scenarios.</li><li>Semantic Understanding: Structured data provides semantic meaning that can aid in the interpretation and generation of natural language. LLMs can leverage structured data to understand the relationships between entities, infer missing information, and generate text that aligns with the underlying semantics of the data.</li></ol><p>Overall, structured data plays a crucial role in enhancing the performance, coherence, and relevance of LLMs and generative AI systems by providing clear signals, context, and semantic understanding. Integrating structured data effectively can lead to more accurate, coherent, and useful outputs from these models across various applications and domains.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=42557c401793" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>