diff options
Diffstat (limited to 'cv/index.typ')
| -rw-r--r-- | cv/index.typ | 103 |
1 files changed, 34 insertions, 69 deletions
diff --git a/cv/index.typ b/cv/index.typ index 63b5c6e..e4f8227 100644 --- a/cv/index.typ +++ b/cv/index.typ @@ -6,58 +6,43 @@ #let email = "jayrupnakawala@gmail.com" #let github = "github.com/CaptainJack2491" #let linkedin = "linkedin.com/in/jayrupnakawala" -#let phone = "+1 (xxx) xxx-xxxx" #let personal-site = "jayrup.me" #show: resume.with( author: name, - // All the lines below are optional. - // For example, if you want to to hide your phone number: - // feel free to comment those lines out and they will not show. location: location, email: email, github: github, linkedin: linkedin, - // phone: phone, personal-site: personal-site, accent-color: "#26428b", font: "New Computer Modern", - paper: "us-letter", + paper: "a4", author-position: center, personal-info-position: center, ) -/* -* Lines that start with == are formatted into section headings -* You can use the specific formatting functions if needed -* The following formatting functions are listed below -* #edu(dates: "", degree: "", gpa: "", institution: "", location: "", consistent: false) -* #work(company: "", dates: "", location: "", title: "") -* #project(dates: "", name: "", role: "", url: "") -* certificates(name: "", issuer: "", url: "", date: "") -* #extracurriculars(activity: "", dates: "") -* There are also the following generic functions that don't apply any formatting -* #generic-two-by-two(top-left: "", top-right: "", bottom-left: "", bottom-right: "") -* #generic-one-by-two(left: "", right: "") -*/ - -/* -== Work Experience +== Research Experience -#work( - title: "Subatomic Shepherd and Caffeine Connoisseur", - location: "Atomville, CA", - company: "Microscopic Circus, Schrodinger's University", - dates: dates-helper(start-date: "May 2024", end-date: "Present"), +#project( + name: "Prime-Grokking: Pre-registered Grokking Study", + url: "git.jayrup.me/c/prime-grokking.git", + dates: dates-helper(start-date: "Aug 2026", end-date: "Present"), ) -- Played God with tiny molecules, making them dance to uncover the secrets of the universe -- Convinced high-performance computers to work overtime without unions, reducing simulation time by 50% -- Wowed a room full of nerds with pretty pictures of invisible things and imaginary findings -*/ +- Tested whether minimal transformers and RNNs can grok next-prime prediction, running the study like production research: every experiment batch pre-registered with commit-locked interpretations before launch. +- Caught two experimental-design blockers (layout invariance, prereg-code drift) via hostile pre-run review; built an idempotent multi-cell sweep runner and batched the evaluation loop *46× faster*. +- Ran a 16-cell PyTorch CUDA/AMP sweep (architecture × weight decay) on a remote GPU node, then extended through 10× more data and a 20× longer horizon — following the negative result rather than tuning it away. +- Published a 3-part writeup at jayrup.me with a public, re-runnable repository (prereg addenda, eval code). +#project( + name: "Stress-Testing the J-space Interpretability Lens", + url: "jayrup.me/blog/jacobian-lens-frequency", + dates: "Jul 2026 — Aug 2026", +) -== Research Experience +- Independently reproduced a frontier interpretability method (Anthropic's July 2026 J-space/J-lens analysis) on a small PyTorch transformer and found its scores track token frequency. +- Designed a matched-frequency control showing contextual predictability remains detectable — framed and published as a control requirement for the method, not a refutation; shared with the mech-interp community. #project( name: "Dissertation: Moral Cover and Watchful Eyes", @@ -65,39 +50,25 @@ role: "", // Dates is optional dates: dates-helper(start-date: "Nov 2025", end-date: "May 2026"), - // URL is also optional - url: "github.com/CaptainJack2491/Dissertation", ) - Architected a sandboxed evaluation framework with a Virtual File System (VFS) to safely execute and monitor autonomous LLM tool-use across multiple providers (Anthropic, OpenAI, Google). - Built a provider abstraction layer for glass-box reasoning extraction and a post-hoc interrogation system to adversarially probe agent intent. - Designed a 3-prong automated judging pipeline (regex, LLM blackbox, LLM glassbox) and ran *960 experimental trials* across exploratory and confirmatory studies. - Quantified a *"Stealth Gap"*: deceptive intent held steady (~62--67%) under high oversight while output-level detectability dropped (47.8%→35.6%), showing output-only monitoring underestimates risk. - +- Code and evaluation harness: github.com/CaptainJack2491/Dissertation == Projects #project( - // role: "Lead Infrastructure Engineer", name: "Hybrid-Cloud Research Infrastructure & Zero-Trust Mesh", - // url: "jayrup.me", dates: dates-helper(start-date: "Jan 2024", end-date: "Present"), ) -- Developed a hybrid-cloud networking mesh using Tailscale/WireGuard to unify a Dell Precision local node (Ubuntu) with 3 distributed VPS instances into a single secure plane. -- Engineered a custom ingress solution using VPS nodes as forward proxies to bypass CGNAT, ensuring secure, high-availability access to internal research tools without port forwarding. -- Orchestrated 15+ containerized services via Docker Compose, implementing automated backups, resource monitoring, and a centralized NVMe-backed storage tier. -- Optimized a headless development workflow (Arch Linux, Neovim, Tmux) with dotfile versioning for consistent, rapid environment replication across heterogeneous hardware. - -#project( - name: "Jayrup.me: Fast Personal Portfolio", - url: "jayrup.me", - dates: "Nov 2025 — Present" -) -- Engineered a dual-interface web server using Nginx to serve context-aware content (HTML vs. Plaintext) based on User-Agent detection (e.g., `curl`, `wget`). -- Architected a Quarto-based publishing pipeline that generates synchronized web and CLI-friendly documentation from a single Markdown source. -- Automated Deployment: Optimized a secure rsync-over-SSH deployment workflow from a local Arch Linux environment to a distributed VPS cluster. -- Focus: Having a light-weight portfolio that is accessible via the terminal +- Operate a multi-node Linux estate (Void, Arch, Ubuntu) — always-on agent laptop, homelab server, remote GPU research node, VPS edge nodes — unified over a Tailscale/WireGuard zero-trust mesh. +- Run 15+ containerized services in production (Nextcloud, media, password vault, document tooling) via Docker Compose with automated backups, resource monitoring, and an NVMe-backed storage tier. +- Solved CGNAT reachability by deploying VPS nodes as forward proxies, restoring secure high-availability access to internal services without port forwarding. +- Built jayrup.me on this stack: Nginx dual-interface serving (HTML for browsers, plaintext for `curl`) with a Quarto pipeline rendering web and CLI docs from one Markdown source. == Education @@ -105,30 +76,24 @@ institution: "University of East London", location: "London, UK", dates: dates-helper(start-date: "Sept 2023", end-date: "May 2026"), - degree: "Bachelor's of Science (Hons) Data Science and Artificial Intelligence — First-Class Honours (94%)", + degree: "BSc (Hons) Data Science & Artificial Intelligence — First-Class Honours (94%)", // Uncomment the line below if you want edu formatting to be consistent with everything else // consistent: true ) -- Relevant Coursework: Math for computing, Data Structure and Algorithm, Networking, Database systems, Big Data analysis, Project Management, Artificial Intelligence - +- Relevant Coursework: Math for Computing, Data Structures & Algorithms, Networking, Databases, Big Data, AI +/* +== Extracurriculars -// #extracurriculars( -// activity: "Science Olympiad Volunteering", -// dates: "Sep 2023 --- Present" -// ) -// - Volunteer and write tests for tournaments, including LA Regionals and SoCal State \@ Caltech - -// #certificates( -// name: "OSCP", -// issuer: "Offensive Security", -// // url: "", -// date: "Oct 2024", -// ) +#extracurriculars( + activity: "GDG on Campus UEL — hands-on dev workshops (developer-portfolio codelab: Antigravity + Firebase)", + dates: "<DATES>", +) +*/ == Skills *Programming*: Python, Bash, SQL\ -*ML/AI*: PyTorch, Transformers, LangChain, scikit-learn\ -*Infrastructure*: Docker, Nginx, Caddy, Tailscale, WireGuard, Quarto \ -*Systems*: Linux (Arch, Ubuntu), Git, rsync, Tmux, Neovim\ +*ML/AI*: PyTorch, Transformers, Hugging Face, llama.cpp (GGUF quant), LangChain, scikit-learn\ +*Infrastructure*: Docker, Docker Compose, Nginx, Caddy, Tailscale, WireGuard, Quarto\ +*Systems*: Linux (Arch, Void, Ubuntu), Git, rsync, Tmux, Neovim\ |