<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Software Engineering on Jaehun's Blog</title><link>https://jaehun.me/tags/software-engineering/</link><description>Recent content in Software Engineering on Jaehun's Blog</description><generator>Hugo</generator><language>ko-kr</language><lastBuildDate>Sat, 19 Sep 2026 08:58:51 +0900</lastBuildDate><atom:link href="https://jaehun.me/tags/software-engineering/index.xml" rel="self" type="application/rss+xml"/><item><title>Same Request, Different Answer: Quantization Amplifies Cache-Induced Divergence in LLM Serving</title><link>https://jaehun.me/posts/same-request-different-answer-quantization-amplifies-cache-induced-divergence-in-llm-serving/</link><pubDate>Mon, 07 Sep 2026 00:00:00 +0900</pubDate><guid>https://jaehun.me/posts/same-request-different-answer-quantization-amplifies-cache-induced-divergence-in-llm-serving/</guid><description>&lt;p&gt;&lt;a&#10; href="https://arxiv.org/abs/2609.04748v1"target="_blank"&#10; class="inline-flex items-center gap-1"&#10; &gt;논문 링크&lt;svg class="h-3 w-3 flex-shrink-0" id="external-link" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"&gt;&lt;path fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M15 3h6v6m-11 5L21 3m-3 10v6a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2V8a2 2 0 0 1 2-2h6"/&gt;&lt;/svg&gt;&#10; &lt;/a&gt;&lt;/p&gt;&#10;&lt;h2 id="같은-요청-다른-답-프리픽스-캐시가-서빙을-비재현적으로-만들고-양자화가-이를-증폭시킨다"&gt;같은 요청, 다른 답: 프리픽스 캐시가 서빙을 비재현적으로 만들고, 양자화가 이를 증폭시킨다&lt;a href="#%ea%b0%99%ec%9d%80-%ec%9a%94%ec%b2%ad-%eb%8b%a4%eb%a5%b8-%eb%8b%b5-%ed%94%84%eb%a6%ac%ed%94%bd%ec%8a%a4-%ec%ba%90%ec%8b%9c%ea%b0%80-%ec%84%9c%eb%b9%99%ec%9d%84-%eb%b9%84%ec%9e%ac%ed%98%84%ec%a0%81%ec%9c%bc%eb%a1%9c-%eb%a7%8c%eb%93%a4%ea%b3%a0-%ec%96%91%ec%9e%90%ed%99%94%ea%b0%80-%ec%9d%b4%eb%a5%bc-%ec%a6%9d%ed%8f%ad%ec%8b%9c%ed%82%a8%eb%8b%a4" class="heading-anchor" aria-label="이 섹션에 대한 링크"&gt;&lt;/a&gt;&lt;/h2&gt;&#10; &lt;blockquote&#10; class="border-primary bg-muted/30 hover:bg-muted/40 my-6 rounded-r-lg border-l-4 py-4 pr-4 pl-6 italic transition-all duration-200 ease-out hover:translate-x-1"&#10; &gt;&#10; &lt;div class="prose prose-sm max-w-none"&gt;&#10; &lt;p&gt;&lt;strong&gt;TL;DR&lt;/strong&gt; — 프리픽스 캐시는 &amp;ldquo;투명한 최적화&amp;quot;로 여겨지지만 투명하지 않다. 캐시를 끄면 동일 워크로드의 반복 실행이 &lt;strong&gt;10개 설정 × 80에피소드 = 800개 전부 비트 단위로 동일&lt;/strong&gt; 하게 재현되는 반면 (근거: Tab. 2), 캐시를 켜면 에이전트의 궤적이 &lt;strong&gt;16비트에서 36.2%, 4비트 양자화에서 75.0%&lt;/strong&gt; 나 바뀐다 (근거: §IV-D). 그런데 평균 정확도는 움직이지 않는다 — 이것은 &amp;ldquo;열화(degradation)&amp;ldquo;가 아니라 &amp;ldquo;불안정성(instability)&amp;ldquo;이며, 둘은 다른 대응을 요구한다 (근거: §IV-F, §V).&lt;/p&gt;</description></item><item><title>AIOpsLab: A Holistic Framework to Evaluate AI Agents for Enabling Autonomous Clouds</title><link>https://jaehun.me/posts/aiopslab-a-holistic-framework-to-evaluate-ai-agents-for-enabling-autonomous-clouds/</link><pubDate>Wed, 02 Apr 2025 00:00:00 +0900</pubDate><guid>https://jaehun.me/posts/aiopslab-a-holistic-framework-to-evaluate-ai-agents-for-enabling-autonomous-clouds/</guid><description>&lt;p&gt;&lt;a&#10; href="https://arxiv.org/abs/2501.06706"target="_blank"&#10; class="inline-flex items-center gap-1"&#10; &gt;논문 링크&lt;svg class="h-3 w-3 flex-shrink-0" id="external-link" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"&gt;&lt;path fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M15 3h6v6m-11 5L21 3m-3 10v6a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2V8a2 2 0 0 1 2-2h6"/&gt;&lt;/svg&gt;&#10; &lt;/a&gt;&lt;/p&gt;&#10;&lt;h2 id="강점과-독창성-핵심-알고리즘-그리고-한계"&gt;강점과 독창성, 핵심 알고리즘, 그리고 한계&lt;a href="#%ea%b0%95%ec%a0%90%ea%b3%bc-%eb%8f%85%ec%b0%bd%ec%84%b1-%ed%95%b5%ec%8b%ac-%ec%95%8c%ea%b3%a0%eb%a6%ac%ec%a6%98-%ea%b7%b8%eb%a6%ac%ea%b3%a0-%ed%95%9c%ea%b3%84" class="heading-anchor" aria-label="이 섹션에 대한 링크"&gt;&lt;/a&gt;&lt;/h2&gt;&#10; &lt;blockquote&#10; class="border-primary bg-muted/30 hover:bg-muted/40 my-6 rounded-r-lg border-l-4 py-4 pr-4 pl-6 italic transition-all duration-200 ease-out hover:translate-x-1"&#10; &gt;&#10; &lt;div class="prose prose-sm max-w-none"&gt;&#10; &lt;p&gt;물어본 것: 이 논문을 아주 자세하게 읽고 논문의 강점과 독창적인 지점을 설명해주고 핵심 알고리즘을 예시 입력을 들어서 전체적인 과정을 설명해줘 추가적으로 논문의 한계점에 대해서도 알려줘&lt;/p&gt;</description></item><item><title>BitsAI-CR: Automated Code Review via LLM in Practice</title><link>https://jaehun.me/posts/bitsai-cr-automated-code-review-via-llm-in-practice/</link><pubDate>Thu, 13 Feb 2025 00:00:00 +0900</pubDate><guid>https://jaehun.me/posts/bitsai-cr-automated-code-review-via-llm-in-practice/</guid><description>&lt;p&gt;&lt;a&#10; href="https://arxiv.org/abs/2501.15134v1"target="_blank"&#10; class="inline-flex items-center gap-1"&#10; &gt;논문 링크&lt;svg class="h-3 w-3 flex-shrink-0" id="external-link" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"&gt;&lt;path fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M15 3h6v6m-11 5L21 3m-3 10v6a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2V8a2 2 0 0 1 2-2h6"/&gt;&lt;/svg&gt;&#10; &lt;/a&gt;&lt;/p&gt;&#10;&lt;h2 id="강점과-독창성-핵심-알고리즘-그리고-한계"&gt;강점과 독창성, 핵심 알고리즘, 그리고 한계&lt;a href="#%ea%b0%95%ec%a0%90%ea%b3%bc-%eb%8f%85%ec%b0%bd%ec%84%b1-%ed%95%b5%ec%8b%ac-%ec%95%8c%ea%b3%a0%eb%a6%ac%ec%a6%98-%ea%b7%b8%eb%a6%ac%ea%b3%a0-%ed%95%9c%ea%b3%84" class="heading-anchor" aria-label="이 섹션에 대한 링크"&gt;&lt;/a&gt;&lt;/h2&gt;&#10; &lt;blockquote&#10; class="border-primary bg-muted/30 hover:bg-muted/40 my-6 rounded-r-lg border-l-4 py-4 pr-4 pl-6 italic transition-all duration-200 ease-out hover:translate-x-1"&#10; &gt;&#10; &lt;div class="prose prose-sm max-w-none"&gt;&#10; &lt;p&gt;물어본 것: 이 논문을 아주 자세하게 읽고 논문의 강점과 독창적인 지점을 설명해주고 핵심 알고리즘을 예시 입력을 들어서 전체적인 과정을 설명해줘 추가적으로 논문의 한계점에 대해서도 알려줘&lt;/p&gt;</description></item><item><title>DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligence</title><link>https://jaehun.me/posts/deepseek-coder-v2-breaking-the-barrier-of-closed-source-models-in-code-intelligence/</link><pubDate>Tue, 11 Feb 2025 00:00:00 +0900</pubDate><guid>https://jaehun.me/posts/deepseek-coder-v2-breaking-the-barrier-of-closed-source-models-in-code-intelligence/</guid><description>&lt;p&gt;&lt;a&#10; href="https://arxiv.org/abs/2406.11931v1"target="_blank"&#10; class="inline-flex items-center gap-1"&#10; &gt;논문 링크&lt;svg class="h-3 w-3 flex-shrink-0" id="external-link" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"&gt;&lt;path fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M15 3h6v6m-11 5L21 3m-3 10v6a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2V8a2 2 0 0 1 2-2h6"/&gt;&lt;/svg&gt;&#10; &lt;/a&gt;&lt;/p&gt;&#10;&lt;h2 id="강점과-독창성-핵심-알고리즘-그리고-한계"&gt;강점과 독창성, 핵심 알고리즘, 그리고 한계&lt;a href="#%ea%b0%95%ec%a0%90%ea%b3%bc-%eb%8f%85%ec%b0%bd%ec%84%b1-%ed%95%b5%ec%8b%ac-%ec%95%8c%ea%b3%a0%eb%a6%ac%ec%a6%98-%ea%b7%b8%eb%a6%ac%ea%b3%a0-%ed%95%9c%ea%b3%84" class="heading-anchor" aria-label="이 섹션에 대한 링크"&gt;&lt;/a&gt;&lt;/h2&gt;&#10; &lt;blockquote&#10; class="border-primary bg-muted/30 hover:bg-muted/40 my-6 rounded-r-lg border-l-4 py-4 pr-4 pl-6 italic transition-all duration-200 ease-out hover:translate-x-1"&#10; &gt;&#10; &lt;div class="prose prose-sm max-w-none"&gt;&#10; &lt;p&gt;물어본 것: 이 논문을 아주 자세하게 읽고 논문의 강점과 독창적인 지점을 설명해주고 핵심 알고리즘을 예시 입력을 들어서 전체적인 과정을 설명해줘 추가적으로 논문의 한계점에 대해서도 알려줘&lt;/p&gt;</description></item><item><title>DeepSeek-Coder: When the Large Language Model Meets Programming - The Rise of Code Intelligence</title><link>https://jaehun.me/posts/deepseek-coder-when-the-large-language-model-meets-programming-the-rise-of-code-intelligence/</link><pubDate>Fri, 07 Feb 2025 00:00:00 +0900</pubDate><guid>https://jaehun.me/posts/deepseek-coder-when-the-large-language-model-meets-programming-the-rise-of-code-intelligence/</guid><description>&lt;p&gt;&lt;a&#10; href="https://arxiv.org/abs/2401.14196v2"target="_blank"&#10; class="inline-flex items-center gap-1"&#10; &gt;논문 링크&lt;svg class="h-3 w-3 flex-shrink-0" id="external-link" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"&gt;&lt;path fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M15 3h6v6m-11 5L21 3m-3 10v6a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2V8a2 2 0 0 1 2-2h6"/&gt;&lt;/svg&gt;&#10; &lt;/a&gt;&lt;/p&gt;&#10;&#10; &lt;blockquote&#10; class="border-primary bg-muted/30 hover:bg-muted/40 my-6 rounded-r-lg border-l-4 py-4 pr-4 pl-6 italic transition-all duration-200 ease-out hover:translate-x-1"&#10; &gt;&#10; &lt;div class="prose prose-sm max-w-none"&gt;&#10; &lt;p&gt;같은 논문을 2025-06-30 에도 &lt;a&#10; href="https://jaehun.me/posts/deepseek-coder-when-the-large-language-model-meets-programming--the-rise-of-code-intelligence/"&gt;한 번 더 정리&lt;/a&gt;해 두었다.&lt;/p&gt;&#10;&#10; &lt;/div&gt;&#10; &lt;/blockquote&gt;&lt;script&gt;&#10; function toggleAlert(alertId) {&#10; const content = document.getElementById(alertId + "-content");&#10; const chevron = document.getElementById(alertId + "-chevron");&#10; const header = content.previousElementSibling;&#10;&#10; if (content.classList.contains("hidden")) {&#10; content.classList.remove("hidden");&#10; chevron.style.transform = "rotate(0deg)";&#10; header.setAttribute("aria-expanded", "true");&#10; } else {&#10; content.classList.add("hidden");&#10; chevron.style.transform = "rotate(-90deg)";&#10; header.setAttribute("aria-expanded", "false");&#10; }&#10; }&#10;&#10; &#10; document.addEventListener("DOMContentLoaded", function () {&#10; const collapsedAlerts = document.querySelectorAll(".alert-content.hidden");&#10; collapsedAlerts.forEach(function (content) {&#10; const alertId = content.id.replace("-content", "");&#10; const chevron = document.getElementById(alertId + "-chevron");&#10; if (chevron) {&#10; chevron.style.transform = "rotate(-90deg)";&#10; }&#10; });&#10; });&#10;&lt;/script&gt;&#10;&lt;h2 id="강점과-독창성-핵심-알고리즘-그리고-한계"&gt;강점과 독창성, 핵심 알고리즘, 그리고 한계&lt;a href="#%ea%b0%95%ec%a0%90%ea%b3%bc-%eb%8f%85%ec%b0%bd%ec%84%b1-%ed%95%b5%ec%8b%ac-%ec%95%8c%ea%b3%a0%eb%a6%ac%ec%a6%98-%ea%b7%b8%eb%a6%ac%ea%b3%a0-%ed%95%9c%ea%b3%84" class="heading-anchor" aria-label="이 섹션에 대한 링크"&gt;&lt;/a&gt;&lt;/h2&gt;&#10; &lt;blockquote&#10; class="border-primary bg-muted/30 hover:bg-muted/40 my-6 rounded-r-lg border-l-4 py-4 pr-4 pl-6 italic transition-all duration-200 ease-out hover:translate-x-1"&#10; &gt;&#10; &lt;div class="prose prose-sm max-w-none"&gt;&#10; &lt;p&gt;물어본 것: 이 논문을 아주 자세하게 읽고 논문의 강점과 독창적인 지점을 설명해주고 핵심 알고리즘을 예시 입력을 들어서 전체적인 과정을 설명해줘 추가적으로 논문의 한계점에 대해서도 알려줘&lt;/p&gt;</description></item><item><title>KernelGPT Enhanced Kernel Fuzzing via Large Language Models</title><link>https://jaehun.me/posts/kernelgpt-enhanced-kernel-fuzzing-via-large-language-models/</link><pubDate>Tue, 05 Nov 2024 00:00:00 +0900</pubDate><guid>https://jaehun.me/posts/kernelgpt-enhanced-kernel-fuzzing-via-large-language-models/</guid><description>&lt;p&gt;논문 : &lt;a&#10; href="https://arxiv.org/abs/2401.00563"target="_blank"&#10; class="inline-flex items-center gap-1"&#10; &gt;https://arxiv.org/abs/2401.00563&lt;svg class="h-3 w-3 flex-shrink-0" id="external-link" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"&gt;&lt;path fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M15 3h6v6m-11 5L21 3m-3 10v6a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2V8a2 2 0 0 1 2-2h6"/&gt;&lt;/svg&gt;&#10; &lt;/a&gt;&lt;/p&gt;&#10;&lt;p&gt;아래글은 Scholar GPT와 대화를 나눈 내용입니다.&lt;/p&gt;&#10;&lt;h2 id="핵심-내용과-강점-알고리즘-한계"&gt;핵심 내용과 강점, 알고리즘, 한계&lt;a href="#%ed%95%b5%ec%8b%ac-%eb%82%b4%ec%9a%a9%ea%b3%bc-%ea%b0%95%ec%a0%90-%ec%95%8c%ea%b3%a0%eb%a6%ac%ec%a6%98-%ed%95%9c%ea%b3%84" class="heading-anchor" aria-label="이 섹션에 대한 링크"&gt;&lt;/a&gt;&lt;/h2&gt;&#10; &lt;blockquote&#10; class="border-primary bg-muted/30 hover:bg-muted/40 my-6 rounded-r-lg border-l-4 py-4 pr-4 pl-6 italic transition-all duration-200 ease-out hover:translate-x-1"&#10; &gt;&#10; &lt;div class="prose prose-sm max-w-none"&gt;&#10; &lt;p&gt;물어본 것: 논문의 핵심 내용과 강점, 알고리즘 설명, 그리고 한계점&lt;/p&gt;</description></item><item><title>Magicoder Empowering Code Generation with OSS-Instruct</title><link>https://jaehun.me/posts/magicoder-empowering-code-generation-with-oss-instruct/</link><pubDate>Tue, 05 Nov 2024 00:00:00 +0900</pubDate><guid>https://jaehun.me/posts/magicoder-empowering-code-generation-with-oss-instruct/</guid><description>&lt;p&gt;논문 : &lt;a&#10; href="https://arxiv.org/abs/2312.02120"target="_blank"&#10; class="inline-flex items-center gap-1"&#10; &gt;https://arxiv.org/abs/2312.02120&lt;svg class="h-3 w-3 flex-shrink-0" id="external-link" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"&gt;&lt;path fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M15 3h6v6m-11 5L21 3m-3 10v6a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2V8a2 2 0 0 1 2-2h6"/&gt;&lt;/svg&gt;&#10; &lt;/a&gt;&lt;/p&gt;&#10;&lt;p&gt;아래글은 Scholar GPT와 대화를 나눈 내용입니다.&lt;/p&gt;&#10;&lt;h2 id="핵심-내용과-강점-알고리즘-한계"&gt;핵심 내용과 강점, 알고리즘, 한계&lt;a href="#%ed%95%b5%ec%8b%ac-%eb%82%b4%ec%9a%a9%ea%b3%bc-%ea%b0%95%ec%a0%90-%ec%95%8c%ea%b3%a0%eb%a6%ac%ec%a6%98-%ed%95%9c%ea%b3%84" class="heading-anchor" aria-label="이 섹션에 대한 링크"&gt;&lt;/a&gt;&lt;/h2&gt;&#10; &lt;blockquote&#10; class="border-primary bg-muted/30 hover:bg-muted/40 my-6 rounded-r-lg border-l-4 py-4 pr-4 pl-6 italic transition-all duration-200 ease-out hover:translate-x-1"&#10; &gt;&#10; &lt;div class="prose prose-sm max-w-none"&gt;&#10; &lt;p&gt;물어본 것: 논문의 핵심 내용과 강점, 알고리즘 설명, 그리고 한계점&lt;/p&gt;</description></item></channel></rss>