<?xml version="1.0" encoding="UTF-8" ?>
<rss version="2.0">
    <channel>
      <title>PUMA Vault</title>
      <link>https://pumacp.github.io/puma-vault</link>
      <description>Últimas 10 notas on PUMA Vault</description>
      <generator>Quartz -- quartz.jzhao.xyz</generator>
      <item>
    <title>Fleeting Note — {{date}} — {{keyword}}</title>
    <link>https://pumacp.github.io/puma-vault/10---Inbox/Fleeting-Notes/Template-Fleeting-Note</link>
    <guid>https://pumacp.github.io/puma-vault/10---Inbox/Fleeting-Notes/Template-Fleeting-Note</guid>
    <description><![CDATA[ Fleeting Note — {{date}} ⚡ Capture now, process later. ]]></description>
    <pubDate>Fri, 01 May 2026 01:47:52 GMT</pubDate>
  </item><item>
    <title>GRAPH_REPORT</title>
    <link>https://pumacp.github.io/puma-vault/graphify-out/GRAPH_REPORT</link>
    <guid>https://pumacp.github.io/puma-vault/graphify-out/GRAPH_REPORT</guid>
    <description><![CDATA[ Graph Report - . (2026-05-01) Corpus Check 448 files · ~249,404 words Verdict: corpus is large enough that graph structure adds value. ]]></description>
    <pubDate>Fri, 01 May 2026 01:46:02 GMT</pubDate>
  </item><item>
    <title>Context Engineering — Designing the LLM Context Window as a System</title>
    <link>https://pumacp.github.io/puma-vault/30---Permanent/31-Concepts/PN-ContextEngineering</link>
    <guid>https://pumacp.github.io/puma-vault/30---Permanent/31-Concepts/PN-ContextEngineering</guid>
    <description><![CDATA[ PN: Context Engineering — Designing the LLM Context Window as a System Core Idea Context Engineering (CE) is the discipline of constructing the entire content that enters an LLM’s context window — including system prompts, retrieved knowledge, tool outputs, conversation history, and structured data ... ]]></description>
    <pubDate>Fri, 01 May 2026 00:00:00 GMT</pubDate>
  </item><item>
    <title>MCP — Model Context Protocol: Architecture, Security, and PUMA Integration</title>
    <link>https://pumacp.github.io/puma-vault/30---Permanent/31-Concepts/PN-MCP-ModelContextProtocol</link>
    <guid>https://pumacp.github.io/puma-vault/30---Permanent/31-Concepts/PN-MCP-ModelContextProtocol</guid>
    <description><![CDATA[ PN: MCP — Model Context Protocol: Architecture, Security, and PUMA Integration Core Idea The Model Context Protocol (MCP) is an open standard (Anthropic, Nov 2024) that defines how LLM agents connect to external tools, data sources, and services through a unified client–server interface. ]]></description>
    <pubDate>Fri, 01 May 2026 00:00:00 GMT</pubDate>
  </item><item>
    <title>RLHF and Constitutional AI — LLM Alignment Training Paradigms</title>
    <link>https://pumacp.github.io/puma-vault/30---Permanent/31-Concepts/PN-RLHF-Constitutional</link>
    <guid>https://pumacp.github.io/puma-vault/30---Permanent/31-Concepts/PN-RLHF-Constitutional</guid>
    <description><![CDATA[ PN: RLHF and Constitutional AI — LLM Alignment Training Paradigms Core Idea RLHF (Reinforcement Learning from Human Feedback) and Constitutional AI (CAI) are the two dominant paradigms for aligning pre-trained LLMs to human values and instructions. ]]></description>
    <pubDate>Fri, 01 May 2026 00:00:00 GMT</pubDate>
  </item><item>
    <title>Transformer Architecture and Mixture of Experts — Technical Reference for PUMA Models</title>
    <link>https://pumacp.github.io/puma-vault/30---Permanent/31-Concepts/PN-Transformer-MoE</link>
    <guid>https://pumacp.github.io/puma-vault/30---Permanent/31-Concepts/PN-Transformer-MoE</guid>
    <description><![CDATA[ PN: Transformer Architecture and Mixture of Experts — Technical Reference for PUMA Models Core Idea The Transformer (Vaswani et al., 2017) is the universal architecture underlying all LLMs used in PUMA. ]]></description>
    <pubDate>Fri, 01 May 2026 00:00:00 GMT</pubDate>
  </item><item>
    <title>Tree of Thoughts — Deliberate Multi-Path Reasoning for LLMs</title>
    <link>https://pumacp.github.io/puma-vault/30---Permanent/31-Concepts/PN-TreeOfThoughts-Deliberate</link>
    <guid>https://pumacp.github.io/puma-vault/30---Permanent/31-Concepts/PN-TreeOfThoughts-Deliberate</guid>
    <description><![CDATA[ PN: Tree of Thoughts — Deliberate Multi-Path Reasoning for LLMs Core Idea Tree of Thoughts (ToT) (Yao et al., NeurIPS 2023) frames LLM problem-solving as a search over a tree of intermediate reasoning steps (“thoughts”). ]]></description>
    <pubDate>Fri, 01 May 2026 00:00:00 GMT</pubDate>
  </item><item>
    <title>Knowledge Coverage Report — PUMA Vault Expansion (April 2026)</title>
    <link>https://pumacp.github.io/puma-vault/30---Permanent/39-Reports/Knowledge-Coverage-Report-2026-04</link>
    <guid>https://pumacp.github.io/puma-vault/30---Permanent/39-Reports/Knowledge-Coverage-Report-2026-04</guid>
    <description><![CDATA[ Knowledge Coverage Report — PUMA Vault Expansion (April 2026) Purpose This report documents the systematic knowledge expansion of the PUMA vault conducted in April 2026. ]]></description>
    <pubDate>Fri, 01 May 2026 00:00:00 GMT</pubDate>
  </item><item>
    <title>🧠 MOC — Methods &amp; Frameworks</title>
    <link>https://pumacp.github.io/puma-vault/80---MOC/81-Topic-Maps/MOC-Methods-Frameworks</link>
    <guid>https://pumacp.github.io/puma-vault/80---MOC/81-Topic-Maps/MOC-Methods-Frameworks</guid>
    <description><![CDATA[ 🧠 MOC — Methods &amp; Frameworks Overview All research, development, and knowledge management methodologies integrated in PUMA and this vault. ]]></description>
    <pubDate>Fri, 01 May 2026 00:00:00 GMT</pubDate>
  </item><item>
    <title>Keshav Three-Pass and MIT AI Lab Three Questions: An Integrated Academic Reading Protocol for PUMA</title>
    <link>https://pumacp.github.io/puma-vault/30---Permanent/33-Frameworks/PN-MIT-Student-Method</link>
    <guid>https://pumacp.github.io/puma-vault/30---Permanent/33-Frameworks/PN-MIT-Student-Method</guid>
    <description><![CDATA[ Keshav Three-Pass and MIT AI Lab Three Questions: An Integrated Academic Reading Protocol for PUMA Atomic Claim Effective reading of research papers requires both systematic coverage (Keshav’s three passes) and generative engagement (MIT AI Lab WP 316’s three questions). ]]></description>
    <pubDate>Wed, 29 Apr 2026 00:00:00 GMT</pubDate>
  </item>
    </channel>
  </rss>