Skip to content
View islamborghini's full-sized avatar
👨‍💻
👨‍💻
20 followers · 24 following

Achievements

Achievement: QuickdrawAchievement: Pull Sharkx2Achievement: YOLO

Achievements

Achievement: QuickdrawAchievement: Pull Sharkx2Achievement: YOLO

Highlights

  • Pro

Block or report islamborghini

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
islamborghini/README.md

Islam Assanov

AI Engineer and Researcher from San Francisco

I study AI and Computer Science at Minerva University. I worked on an independent AI research on red-teaming, evals, and benchmarks. I have previously built my benchmark (check GameBoy Eval), replicated PAIR (check my blog article). Previously, I had 2 internships in SF and Taipei, won Cal Hacks, Cal Hack For Impact, Seoul AI Hackathon.

I am looking to colaborate on AI research and coding projects. I am particularly interested in infra, AI safety, and benchmarks. If you have an idea in mind, share it and we can make it happen.

Check my website: islamassanov.com

To reach out:

My skills

C++ CSS3 HTML5 JavaScript TypeScript Python PHP

Node.js Express.js React TailwindCSS Next.js

PostgreSQL MySQL MongoDB

Docker Google Cloud AWS

Pinned Loading

  1. autonomous-research-for-proteinGym autonomous-research-for-proteinGym Public

    An autonomous research system that beats the current SOTA in ProteinGym DMS-Indels

    Jupyter Notebook

  2. gameboy-eval gameboy-eval Public

    Can an LLM write a working Game Boy emulator? An open, deterministic benchmark that grades each model's emulator against a SameBoy reference

    Python

  3. cogni2 cogni2 Public

    A token-efficient context layer for coding agents. Cogni cuts the tokens, and the cost, an agent spends per task without lowering how often it succeeds.

    Go

  4. clowdy clowdy Public

    Simplified serverless platform for Python. Editor + deploy + invoke + gateway + versions + managed DB, all from one dashboard

    Python

  5. livecaps livecaps Public

    LiveCaps is a browser-based application that provides real-time speech transcription and translation

    TypeScript 2

  6. JailbreakingLLMs JailbreakingLLMs Public

    Forked from patrickrchao/JailbreakingLLMs

    Replication of study (Chao et. al., 2023) on jailbreaking LLMs using PAIR algorithm

    Python