<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: vicentwu</title><link>https://news.ycombinator.com/user?id=vicentwu</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 23 Apr 2026 12:40:37 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=vicentwu" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by vicentwu in "Show HN: How I topped the HuggingFace open LLM leaderboard on two gaming GPUs"]]></title><description><![CDATA[
<p>Good read.</p>
]]></description><pubDate>Wed, 11 Mar 2026 03:45:34 +0000</pubDate><link>https://news.ycombinator.com/item?id=47331543</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=47331543</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47331543</guid></item><item><title><![CDATA[New comment by vicentwu in "Magical Mushroom – Europe's first industrial-scale mycelium packaging producer"]]></title><description><![CDATA[
<p>cool</p>
]]></description><pubDate>Mon, 23 Feb 2026 14:03:05 +0000</pubDate><link>https://news.ycombinator.com/item?id=47122459</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=47122459</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47122459</guid></item><item><title><![CDATA[New comment by vicentwu in "Warcraft III Peon Voice Notifications for Claude Code"]]></title><description><![CDATA[
<p>Genuis!</p>
]]></description><pubDate>Thu, 12 Feb 2026 12:54:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=46988218</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=46988218</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46988218</guid></item><item><title><![CDATA[New comment by vicentwu in "Understanding Neural Network, Visually"]]></title><description><![CDATA[
<p>I like the CRT-like filter effect.</p>
]]></description><pubDate>Sat, 07 Feb 2026 05:54:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=46921677</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=46921677</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46921677</guid></item><item><title><![CDATA[New comment by vicentwu in "Solving a million-step LLM task with zero errors"]]></title><description><![CDATA[
<p>Hmm...
The key is to successfully decompose a big, hard problem into easier atomic sub-problems. However, the decomposition process itself is difficult, and this paper is not about that. They decompose a task using a human-written prompt.</p>
]]></description><pubDate>Wed, 19 Nov 2025 09:43:45 +0000</pubDate><link>https://news.ycombinator.com/item?id=45977646</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=45977646</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45977646</guid></item><item><title><![CDATA[New comment by vicentwu in "AMD signs AI chip-supply deal with OpenAI, gives it option to take a 10% stake"]]></title><description><![CDATA[
<p>wow....</p>
]]></description><pubDate>Mon, 06 Oct 2025 14:58:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=45492132</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=45492132</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45492132</guid></item><item><title><![CDATA[New comment by vicentwu in "The Beer Can (2023)"]]></title><description><![CDATA[
<p>I like the vibe.</p>
]]></description><pubDate>Fri, 03 Oct 2025 13:06:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=45462545</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=45462545</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45462545</guid></item><item><title><![CDATA[New comment by vicentwu in "Claude 4 System Card"]]></title><description><![CDATA[
<p>Prompts are really an interesting way of programming, and we can actually express logic containing abstract adjectives like ‘happy’ and ‘unsatisfied’ in a somewhat arbitrary way.</p>
]]></description><pubDate>Mon, 26 May 2025 13:30:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=44097276</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=44097276</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44097276</guid></item><item><title><![CDATA[New comment by vicentwu in "Richard Sutton and Andrew Barto Win 2024 Turing Award"]]></title><description><![CDATA[
<p>Great!</p>
]]></description><pubDate>Wed, 05 Mar 2025 14:16:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=43266671</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=43266671</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43266671</guid></item><item><title><![CDATA[New comment by vicentwu in "Life is more than an engineering problem"]]></title><description><![CDATA[
<p>"As an analogy, imagine that you could put your dog or cat into hibernate mode whenever you left on a trip. Your dog or cat might not notice, but even if they did, they might not mind. Now imagine that you could put your child into hibernate mode whenever you were too busy to spend time with them. Your child would absolutely notice, and even if you told them it was for their own good, they would make certain inferences about how much you valued them. That’s the situation the human characters in the story find themselves in."
Fascinating.</p>
]]></description><pubDate>Mon, 03 Feb 2025 08:49:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=42916243</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=42916243</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42916243</guid></item><item><title><![CDATA[New comment by vicentwu in "Nvidia’s $589B DeepSeek rout"]]></title><description><![CDATA[
<p>RL doesn't need that much static data, it needs a lot of "good" tasks/challenges and computation.</p>
]]></description><pubDate>Mon, 27 Jan 2025 13:59:39 +0000</pubDate><link>https://news.ycombinator.com/item?id=42841208</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=42841208</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42841208</guid></item><item><title><![CDATA[New comment by vicentwu in "OpenAI O3 breakthrough high score on ARC-AGI-PUB"]]></title><description><![CDATA[
<p>Trained with 300 raw pairs directly from the ARC training set without using any data augmentation process, such as generating many more pairs with some kind of ARC generator? That's amazing.</p>
]]></description><pubDate>Sun, 22 Dec 2024 08:57:34 +0000</pubDate><link>https://news.ycombinator.com/item?id=42485198</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=42485198</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42485198</guid></item><item><title><![CDATA[New comment by vicentwu in "OpenAI O3 breakthrough high score on ARC-AGI-PUB"]]></title><description><![CDATA[
<p>"Note on "tuned": OpenAI shared they trained the o3 we tested on 75% of the Public Training set. They have not shared more details. We have not yet tested the ARC-untrained model to understand how much of the performance is due to ARC-AGI data."<p>Really want to see the number of training pairs needed to achieve this socre. If it only takes a few pairs, say 100 pairs, I would say it is amazing!</p>
]]></description><pubDate>Sat, 21 Dec 2024 06:46:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=42477937</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=42477937</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42477937</guid></item><item><title><![CDATA[New comment by vicentwu in "Intel announces Arc B-series "Battlemage" discrete graphics with Linux support"]]></title><description><![CDATA[
<p>Off the topoc. I think, in the long-term , inference should be done along with some kind of training.</p>
]]></description><pubDate>Wed, 04 Dec 2024 02:23:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=42313939</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=42313939</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42313939</guid></item><item><title><![CDATA[New comment by vicentwu in "Mira Murati leaves OpenAI"]]></title><description><![CDATA[
<p>Past efforts leds to today's products. We need to wait to see the real imapct on the ability to ship.</p>
]]></description><pubDate>Thu, 26 Sep 2024 02:16:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=41653860</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=41653860</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41653860</guid></item><item><title><![CDATA[New comment by vicentwu in "SAM 2: Segment Anything in Images and Videos"]]></title><description><![CDATA[
<p>It's amazing!</p>
]]></description><pubDate>Tue, 30 Jul 2024 01:53:46 +0000</pubDate><link>https://news.ycombinator.com/item?id=41105431</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=41105431</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41105431</guid></item><item><title><![CDATA[New comment by vicentwu in "New algorithm discovers language just by watching videos"]]></title><description><![CDATA[
<p>Fasciating. Language is a type of action evoloved for information exchaging, which maps latent "video", "audio" and "thoughts" into "sentences" and vice versa.</p>
]]></description><pubDate>Sat, 15 Jun 2024 14:21:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=40689993</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=40689993</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=40689993</guid></item><item><title><![CDATA[New comment by vicentwu in "Real-time image editing using latent consistency models"]]></title><description><![CDATA[
<p>Cool! The real-time feedback will have enormous ramifications on the art creation workflows.</p>
]]></description><pubDate>Sat, 11 Nov 2023 07:04:42 +0000</pubDate><link>https://news.ycombinator.com/item?id=38228171</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=38228171</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=38228171</guid></item><item><title><![CDATA[New comment by vicentwu in "How is ChatGPT's behavior changing over time?"]]></title><description><![CDATA[
<p>Asking models to do math is kind of an effecitve way to measure their capabilities, especially in reasoning and abstraction, which are quite important for problem solving.</p>
]]></description><pubDate>Wed, 19 Jul 2023 13:29:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=36785985</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=36785985</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36785985</guid></item><item><title><![CDATA[New comment by vicentwu in "Language Is Not All You Need: Aligning Perception with Language Models"]]></title><description><![CDATA[
<p>The chain-of-thought works quite like the "System 2" introduced in <<Thinking, Fast and Slow>>, which is slower, more deliberative, and more logical.</p>
]]></description><pubDate>Tue, 28 Feb 2023 14:45:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=34969810</link><dc:creator>vicentwu</dc:creator><comments>https://news.ycombinator.com/item?id=34969810</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=34969810</guid></item></channel></rss>