<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
	xmlns:slash="http://purl.org/rss/1.0/modules/slash/"
		>

<channel>
	<title>UniResearch Community | Gary Robinson | Activity</title>
	<link>https://community.uniresearch.ai/members/garyrobinson/activity/</link>
	<atom:link href="https://community.uniresearch.ai/members/garyrobinson/activity/feed/" rel="self" type="application/rss+xml" />
	<description>Activity feed for Gary Robinson.</description>
	<lastBuildDate>Fri, 11 Sep 2026 11:08:05 +0800</lastBuildDate>
	<generator>https://buddypress.org/?v=3.4.4</generator>
	<language>en-US</language>
	<ttl>30</ttl>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>2</sy:updateFrequency>
		
								<item>
				<guid isPermaLink="false">03ab805856bc74b388f755fdafa760d0</guid>
				<title>Gary Robinson posted a new activity comment</title>
				<link>https://community.uniresearch.ai/news-feed/p/1007/#acomment-1031</link>
				<pubDate>Wed, 09 Sep 2026 13:47:54 +0800</pubDate>

									<content:encoded><![CDATA[<p>Document why you chose each ablated component in lab notes before running experiments.</p>
				<strong>In reply to</strong> -
					<a href="https://community.uniresearch.ai/members/xiuhuawei/" data-bb-hp-profile="359" rel="nofollow">Xiuhua Wei</a> posted an update <p><strong>Method Help</strong></p>
<p>Poorly‑designed ablation studies waste computing resources and deliver limited insight. Many new researchers struggle to decide which components should be ablated. What [&hellip;]</p>					]]></content:encoded>
				
				
							</item>
					<item>
				<guid isPermaLink="false">34222ac91752f9a137b3175f2c247599</guid>
				<title>Gary Robinson posted a new activity comment</title>
				<link>https://community.uniresearch.ai/news-feed/p/542/#acomment-552</link>
				<pubDate>Sun, 23 Aug 2026 12:29:24 +0800</pubDate>

									<content:encoded><![CDATA[<p>Start with small, well‑defined sub‑problems instead of chasing huge ambitious research questions at first.</p>
				<strong>In reply to</strong> -
					<a href="https://community.uniresearch.ai/members/juanma/" data-bb-hp-profile="329" rel="nofollow">Juan Ma</a> posted an update Many graduate students excel at exams but struggle when they first face fully open‑ended research problems. What practical advice would you give for this tough transition phase?					]]></content:encoded>
				
				
							</item>
		
	</channel>
</rss>
		