<?xml version='1.0' encoding='utf-8'?>
<rss xmlns:atom="http://www.w3.org/2005/Atom" version="2.0">
  <channel>
    <title>Between Signals — Suvra Nath</title>
    <link>https://suvranath.com/blog.html</link>
    <description>Notes on neuroscience, computation, books, technology, and society.</description>
    <language>en</language>
    <lastBuildDate>Sun, 11 Oct 2026 09:07:48 GMT</lastBuildDate>
    <atom:link href="https://suvranath.com/feed.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Confidently Wrong: How to Catch AI Mistakes Before They Reach Your Bench Science</title>
      <link>https://medium.com/@suvranath047/confidently-wrong-how-to-catch-ai-mistakes-before-they-reach-your-bench-science-403fc1685217?sharedUserId=suvranath047</link>
      <description>A reproducible, evidence-based evaluation strategy formulation demo workflow for biomedical research.</description>
      <category>Neurons and Numbers</category>
      <guid isPermaLink="true">https://medium.com/@suvranath047/confidently-wrong-how-to-catch-ai-mistakes-before-they-reach-your-bench-science-403fc1685217</guid>
      <pubDate>Thu, 08 Oct 2026 22:00:00 GMT</pubDate>
    </item>
    <item>
      <title>Spec-Driven Development with AI Agents: Spec Kit vs OpenSpec vs BMAD (Plus a Hands-On OpenSpec Tutorial)</title>
      <link>https://medium.com/@suvranath047/7b6997074582</link>
      <description>Three tools, one big idea: give your AI coding agent a memory. Here’s how they compare, and how to try the lightest one in under an hour.</description>
      <category>Neurons and Numbers</category>
      <guid isPermaLink="true">https://medium.com/@suvranath047/7b6997074582</guid>
      <pubDate>Mon, 05 Oct 2026 22:00:00 GMT</pubDate>
    </item>
    <item>
      <title>Debugging LLMs Without Guesswork: A Practical Langfuse Tutorial</title>
      <link>https://medium.com/@suvranath047/debugging-llms-without-guesswork-a-practical-langfuse-tutorial-03cb26358f5d</link>
      <description>A practical journey from opaque LLM outputs to transparent, traceable, and observable AI workflows with Langfuse.</description>
      <category>Neurons and Numbers</category>
      <guid isPermaLink="true">https://medium.com/@suvranath047/debugging-llms-without-guesswork-a-practical-langfuse-tutorial-03cb26358f5d</guid>
      <pubDate>Fri, 02 Oct 2026 22:00:00 GMT</pubDate>
    </item>
    <item>
      <title>RAG: Teaching AI to Look Things Up Before It Speaks</title>
      <link>https://suvranath.com/posts/perspective/rag-teaching-ai-to-look-things-up.html</link>
      <description>From TF-IDF and cosine similarity to biological knowledge retrieval and why this could become surprisingly important for drug discovery.</description>
      <category>Neurons and Numbers</category>
      <guid isPermaLink="false">https://suvrazastrovision.github.io/posts/perspective/rag-teaching-ai-to-look-things-up.html</guid>
      <pubDate>Wed, 30 Sep 2026 22:00:00 GMT</pubDate>
    </item>
    <item>
      <title>Mac mini M6: could this tiny machine change how scientists compute?</title>
      <link>https://suvranath.com/posts/perspective/mac-mini-m6-review.html</link>
      <description>A computational scientist’s perspective on local AI, everyday research workflows, and the limits of a compact workstation.</description>
      <category>Perspective</category>
      <guid isPermaLink="false">https://suvrazastrovision.github.io/posts/perspective/mac-mini-m6-review.html</guid>
      <pubDate>Fri, 25 Sep 2026 22:00:00 GMT</pubDate>
    </item>
    <item>
      <title>Empire of AI: a review of power and governance</title>
      <link>https://suvranath.com/posts/perspective/empire-of-ai-review.html</link>
      <description>A personal review of Karen Hao’s reporting on AI, and where the empire metaphor falls short.</description>
      <category>Perspective</category>
      <guid isPermaLink="false">https://suvrazastrovision.github.io/posts/perspective/empire-of-ai-review.html</guid>
      <pubDate>Fri, 21 Aug 2026 22:00:00 GMT</pubDate>
    </item>
    <item>
      <title>Depression might not be a "chemical imbalance" - it might be a stalled construction project in your brain</title>
      <link>https://suvranath.com/posts/neurons-and-numbers/depression-neurogenesis.html</link>
      <description>How stalled neuron growth in the brain may offer a new perspective on depression.</description>
      <category>Neurons and Numbers</category>
      <guid isPermaLink="false">https://suvrazastrovision.github.io/posts/neurons-and-numbers/depression-neurogenesis.html</guid>
      <pubDate>Tue, 18 Aug 2026 08:30:00 GMT</pubDate>
    </item>
  </channel>
</rss>