Skip to content
View YuvGev00's full-sized avatar

Achievements

Achievement: Pull Shark

Achievements

Achievement: Pull Shark

Block or report YuvGev00

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.

Content in all repositories owned by your account will be closed.
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
YuvGev00/README.md

Hi, I'm Yuval

Final-semester Computer Science student at Reichman University (graduating February 2027), looking for junior AI/ML and software engineering roles in Israel. I like building systems I can measure: frozen test splits, CI on every change, and an honest list of what didn't work.

Site: yuvgev.com · Email: yuvgev1@gmail.com

Projects

  • taintwall: a runtime firewall for tool-using LLM agents. A taint-tracking policy blocks side-effecting tool calls (send money, send email) whose arguments came only from untrusted tool output. On 461 held-out AgentDojo attacks against a live model that falls for them, it blocks as many as the ProtectAI detector (55.5% → 13.7%) while keeping 72% of benign tasks working versus 44%. Python, FastAPI, Docker
  • Hebrew ToolBench: a Hebrew tool-calling benchmark: 800 items over 52 Israeli-domain tools, scored by a deterministic AST scorer, with 9 models evaluated. The Hebrew calendar is the hardest part for every model, and the "wrong language for a canonical value" failure appears only in the smallest model. Python
  • mini-engine: an LLM inference engine written from scratch in PyTorch, with a paged KV cache, continuous batching, prefix caching and n-gram speculative decoding. Its greedy output is token-identical to Hugging Face, and it is benchmarked against vLLM on an L4: speculative decoding gives 4.5–4.7× on copy-heavy work. Python, PyTorch
  • iHelp: a reverse help marketplace where verified neighbors compete to help (live). Every permission rule is enforced inside PostgreSQL with row-level security, plus 65 automated tests. Next.js, TypeScript, Supabase
  • SubScan: a Hebrew web app that finds forgotten subscriptions in the statements you export from Israeli card issuers and banks, and drafts the cancellation letters. In beta; the code is private. TypeScript
  • penguin-forge: a real-time multiplayer world on an HTTP/WebSocket framework I hand-wrote on Node's raw net module, with zero dependencies. JavaScript, Node.js

Open source

Pinned Loading

  1. taintwall taintwall Public

    Runtime firewall for tool-using LLM agents: ML injection detector + taint-tracking policy, measured on AgentDojo

    Python

  2. mini-engine mini-engine Public

    From-scratch LLM inference engine: paged KV cache, continuous batching, prefix caching, n-gram speculative decoding; token-identical to HF, benchmarked vs vLLM

    Python

  3. ihelp ihelp Public

    Reverse help-marketplace: post one request, verified neighbors compete to help. Next.js + Supabase, permissions enforced in Postgres RLS.

    TypeScript

  4. obsidian-study-guide obsidian-study-guide Public

    Claude Code skill: turn course material (lecture PDFs, study guides, notebooks) into a deeply interlinked Obsidian study vault — atomic term/formula notes, teaching-walkthrough lectures, visuals, p…

    Python 1

  5. penguin-forge penguin-forge Public

    Real-time multiplayer penguin world + the zero-dependency HTTP/WebSocket framework that runs it, hand-written on Node's raw net module.

    JavaScript

  6. hebrew-toolbench hebrew-toolbench Public

    Hebrew tool-calling benchmark: 800 native-Hebrew items, 52 Israeli-domain tools, deterministic AST scorer, 9 models

    Python