<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Solving the Rubik's Cube Without Human Knowledge · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/1805.07470/</link><description>Collected HN comments and submissions linking this paper.</description><language>en</language><lastBuildDate>Mon, 02 Nov 2026 11:16:22 GMT</lastBuildDate><atom:link href="https://hackerpaper.kaipfister.de/p/1805.07470/feed.xml" rel="self" type="application/rss+xml" /><item><title>Solving the Rubik's Cube Without Human Knowledge — mentioned by MrQuincle</title><link>https://news.ycombinator.com/item?id=21307360</link><guid isPermaLink="false">urn:hackerpapers:mention:1805.07470:21307360</guid><pubDate>Sun, 20 Oct 2019 21:42:39 GMT</pubDate><description>&lt;p&gt;Solving it analytically seems to use deep nets and Monte Carlo tree search [1]. Some additional structure to deep learning and you&amp;#x27;re done. Manipulation is the harder problem. It&amp;#x27;s like complaining that a juggler can&amp;#x27;t count to three. [1] https://arxiv.org/abs/1805.07470&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1805.07470/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.AI</category></item><item><title>Solving the Rubik's Cube Without Human Knowledge — mentioned by sytelus</title><link>https://news.ycombinator.com/item?id=21264965</link><guid isPermaLink="false">urn:hackerpapers:mention:1805.07470:21264965</guid><pubDate>Tue, 15 Oct 2019 23:02:08 GMT</pubDate><description>&lt;p&gt;Many caviates but impressive progress in manipulation, especially sim2real: - Only 20% attempts successful on hardest configs with 26+ moves - Solving steps are not generated by RL (but could be[1]) - Cube is modified internally to transmit additional state via bluetooth - Highly calibrated and fine tuned environment+MuJoCo based sim to match simulation to reality as much as possible - Open AI Five algorithm is pretty much reused as-is - Cumulative training time = 13 thousand years, same order of magnitude as the 40 thousand years - 32+64 V100 GPUs per training cycle [1]…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1805.07470/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.AI</category></item><item><title>Solving the Rubik's Cube Without Human Knowledge — HN submission</title><link>https://news.ycombinator.com/item?id=17633061</link><guid isPermaLink="false">urn:hackerpapers:submission:1805.07470:17633061</guid><pubDate>Sat, 28 Jul 2018 13:29:48 GMT</pubDate><description>&lt;p&gt;Solving the Rubik&amp;#x27;s Cube with Deep Reinforcement Learning&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;1 point; 0 discussion replies on HN. &lt;a href="https://hackerpaper.kaipfister.de/p/1805.07470/"&gt;Paper and source history&lt;/a&gt;. A submission is not an endorsement.&lt;/p&gt;</description><category>cs.AI</category></item><item><title>Solving the Rubik's Cube Without Human Knowledge — mentioned by partycoder</title><link>https://news.ycombinator.com/item?id=17328901</link><guid isPermaLink="false">urn:hackerpapers:mention:1805.07470:17328901</guid><pubDate>Sat, 16 Jun 2018 19:31:25 GMT</pubDate><description>&lt;p&gt;https://arxiv.org/pdf/1805.07470.pdf Results are compared with known optimal algorithms for the Rubik&amp;#x27;s cube: Kociemba&amp;#x27;s algorithm, Korf&amp;#x27;s algorithm. https://en.wikipedia.org/wiki/Optimal_solutions_for_Rubik%27...&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1805.07470/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.AI</category></item><item><title>Solving the Rubik's Cube Without Human Knowledge — HN submission</title><link>https://news.ycombinator.com/item?id=17320007</link><guid isPermaLink="false">urn:hackerpapers:submission:1805.07470:17320007</guid><pubDate>Fri, 15 Jun 2018 14:28:03 GMT</pubDate><description>&lt;p&gt;Solving the Rubik&amp;#x27;s Cube Without Human Knowledge&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;4 points; 0 discussion replies on HN. &lt;a href="https://hackerpaper.kaipfister.de/p/1805.07470/"&gt;Paper and source history&lt;/a&gt;. A submission is not an endorsement.&lt;/p&gt;</description><category>cs.AI</category></item><item><title>Solving the Rubik's Cube Without Human Knowledge — HN submission</title><link>https://news.ycombinator.com/item?id=17131293</link><guid isPermaLink="false">urn:hackerpapers:submission:1805.07470:17131293</guid><pubDate>Wed, 23 May 2018 03:27:57 GMT</pubDate><description>&lt;p&gt;Solving the Rubik&amp;#x27;s Cube Without Human Knowledge&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;1 point; 0 discussion replies on HN. &lt;a href="https://hackerpaper.kaipfister.de/p/1805.07470/"&gt;Paper and source history&lt;/a&gt;. A submission is not an endorsement.&lt;/p&gt;</description><category>cs.AI</category></item></channel></rss>
