From e7dc4befde5bef83d198a631926174bae612f866 Mon Sep 17 00:00:00 2001 From: Void Agent Date: Tue, 1 Sep 2026 22:32:49 +0100 Subject: fix: titles, nav wiring, typos, server_tokens (preview branch — does not deploy) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- blog/how-this-site-works.qmd | 5 +++++ blog/index.qmd | 5 +++++ blog/jacobian-lens-frequency.qmd | 5 +++++ blog/prime-grokking-1.qmd | 5 +++++ blog/prime-grokking-2.qmd | 5 +++++ blog/prime-grokking-3.qmd | 5 +++++ index.qmd | 10 +++++++++- nginx.conf | 3 +++ projects/dis/index.qmd | 7 ++++++- projects/index.qmd | 11 ++++++++--- 10 files changed, 56 insertions(+), 5 deletions(-) diff --git a/blog/how-this-site-works.qmd b/blog/how-this-site-works.qmd index d142a59..45731e7 100644 --- a/blog/how-this-site-works.qmd +++ b/blog/how-this-site-works.qmd @@ -3,6 +3,11 @@ title: "How This Site Works" date: "2026-08-05" --- + + + ## WHY I wanted publishing to be easy, and I wanted this site to diff --git a/blog/index.qmd b/blog/index.qmd index 35710e2..2bf6775 100644 --- a/blog/index.qmd +++ b/blog/index.qmd @@ -2,6 +2,11 @@ title: "Jayrup Nakawala | Blog" --- + + + RANDOM THOUGHTS, EXPERIMENTS, AND SYSTEMS NOTES -------------------------------------------------- diff --git a/blog/jacobian-lens-frequency.qmd b/blog/jacobian-lens-frequency.qmd index 6a81809..e7a516d 100644 --- a/blog/jacobian-lens-frequency.qmd +++ b/blog/jacobian-lens-frequency.qmd @@ -3,6 +3,11 @@ title: "What the Jacobian Lens Measures" date: "2026-08-07" --- + + + *A small replication of Anthropic's J-lens, the token-frequency confound I found, and the bug I almost published.* *This is a story about trying to look inside a language model. I found something diff --git a/blog/prime-grokking-1.qmd b/blog/prime-grokking-1.qmd index 5898131..e6ac85a 100644 --- a/blog/prime-grokking-1.qmd +++ b/blog/prime-grokking-1.qmd @@ -3,6 +3,11 @@ title: "Can a Tiny Network Learn What a Prime Is?" date: "2026-08-14" --- + + + *I trained two tiny neural networks to compute the next prime after a number, to see whether either would discover the algorithm instead of memorizing the answers. Neither did. The way they failed is more interesting than the failure.* diff --git a/blog/prime-grokking-2.qmd b/blog/prime-grokking-2.qmd index f97dcfc..ed23fe6 100644 --- a/blog/prime-grokking-2.qmd +++ b/blog/prime-grokking-2.qmd @@ -3,6 +3,11 @@ title: "Still No Grokking: The Weight-Decay Sweep" date: "2026-08-16" --- + + + *Part 1 found no grokking at weight decay 1.0, and the seed replication confirmed it. So I did the obvious thing: swept the weight decay across five orders of magnitude and watched. Still no grokking. The way it diff --git a/blog/prime-grokking-3.qmd b/blog/prime-grokking-3.qmd index dab2e37..3ff0e02 100644 --- a/blog/prime-grokking-3.qmd +++ b/blog/prime-grokking-3.qmd @@ -3,6 +3,11 @@ title: "Where Exactly Is the Wall?" date: "2026-08-17" --- + + + *The first three phases established that a tiny network cannot grok the next-prime function, and that weight decay doesn't help. This post is the forensic part: four pre-registered diagnostics to find out where the wall diff --git a/index.qmd b/index.qmd index 363c5b2..d06ad06 100644 --- a/index.qmd +++ b/index.qmd @@ -1,3 +1,9 @@ +--- +format: + html: + title: "Jayrup Nakawala" +--- + JAYRUP NAKAWALA -------------------------------------------------- Independent AI-safety researcher @@ -25,6 +31,7 @@ inspected, and experiments that fail loudly. -------------------------------------------------- + -[Note] If you are on windows use curl.exe +[Note] If you are on Windows use curl.exe -------------------------------------------------- diff --git a/nginx.conf b/nginx.conf index c8668c0..d4517f5 100644 --- a/nginx.conf +++ b/nginx.conf @@ -14,6 +14,9 @@ server { server_name jayrup.me; root /usr/share/nginx/html; + # Don't leak nginx version on error pages / server header + server_tokens off; + # Serve PDFs directly location ~* \.pdf$ { add_header Content-Type application/pdf; diff --git a/projects/dis/index.qmd b/projects/dis/index.qmd index ef62874..b53f9b6 100644 --- a/projects/dis/index.qmd +++ b/projects/dis/index.qmd @@ -1,7 +1,12 @@ --- -title: "Dissertation" +title: "Jayrup Nakawala | Dissertation" --- + + + ## MORAL COVER AND WATCHFUL EYES: THE LIMITS OF IN-CONTEXT INTERVENTIONS ON LLM DECEPTION Jayrup Nakawala \ diff --git a/projects/index.qmd b/projects/index.qmd index 840c053..9b44f24 100644 --- a/projects/index.qmd +++ b/projects/index.qmd @@ -2,8 +2,13 @@ title: "Jayrup Nakawala | Projects" --- + + + ## [1] DECEPTIVE ALIGNMENT IN AUTONOMOUS LLM AGENTS \ -[Github](https://github.com/CaptainJack2491/dissertation) | [Full Summary](https://jayrup.me/dissertation) \ +[GitHub](https://github.com/CaptainJack2491/dissertation) | [Full Summary](https://jayrup.me/dissertation) \ Undergraduate Dissertation (Final Year) Studied whether in-context oversight and moral framing @@ -42,13 +47,13 @@ rewards website and automated collecting rewards using a script attached to a cron job. Worked with JWTs and jq to parse the response. First -used curl and jq for prototyping and then implimented +used curl and jq for prototyping and then implemented in python for production. -------------------------------------------------- ## [4] THIS WEBSITE \ -[Github](https://github.com/CaptainJack2491/website) +[GitHub](https://github.com/CaptainJack2491/website) A personal website designed to be read from the terminal. Every page renders as both HTML and plain -- cgit v1.2.3