<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Research | mmrahneshin</title><link>https://mmrahneshin.github.io/tags/research/</link><atom:link href="https://mmrahneshin.github.io/tags/research/index.xml" rel="self" type="application/rss+xml"/><description>Research</description><generator>Hugo Blox Builder (https://hugoblox.com)</generator><language>en-us</language><lastBuildDate>Fri, 05 Jan 2024 00:00:00 +0000</lastBuildDate><image><url>https://mmrahneshin.github.io/media/icon_hu13105007374921565109.png</url><title>Research</title><link>https://mmrahneshin.github.io/tags/research/</link></image><item><title>Presentation: Competition-Level Code Generation with AlphaCode by DeepMind</title><link>https://mmrahneshin.github.io/project/presentation/</link><pubDate>Fri, 05 Jan 2024 00:00:00 +0000</pubDate><guid>https://mmrahneshin.github.io/project/presentation/</guid><description>&lt;p>Presented a technical overview of &lt;mark>&lt;a href="https://deepmind.google/discover/blog/competitive-programming-with-alphacode/" target="_blank" rel="noopener">AlphaCode&lt;/a>&lt;/mark>, DeepMind&amp;rsquo;s LLM for competitive programming, as part of the graduate-level &lt;em>Neural Networks and Deep Learning course&lt;/em>. Covered its architecture, sampling strategies, and evaluation process.&lt;/p>
&lt;p>Discussed AlphaCode’s use of large-scale code generation and filtering techniques to tackle programming problems, along with its positioning among human competitors in programming contests.&lt;/p></description></item></channel></rss>