I'm a data scientist and ML researcher pursuing a research M.Sc. at the Hebrew University of Jerusalem, with a physics foundation (B.Sc. through Bar-Ilan's gifted-in-math program) and years of applied quantitative work — I led an Operations Research & Data Science team in Unit 8200 and built statistical models in the IAF's Operations Research Branch, work recognized with the Israel Defense Prize.
I build ML systems with careful evaluation, reproducible experiments, and statistically grounded analysis. The through-line is trustworthy results — I try not to trust a number until I understand how it was produced.
- Formal reasoning — fine-tuning LLMs to prove theorems in Lean 4 with GRPO, rewarded by the Lean kernel's accept-or-reject, measured against goal-blind frequency baselines.
- Rigorous evaluation — chronological splits over shuffled ones, frequency-table baselines, artifact-scored reproductions, closed forms over Monte-Carlo intuition.
- LLMs & agentic pipelines — evaluation harnesses that score what a run actually produced on disk, not what the agent claims.
- Statistics & operations research — experimental design, uncertainty, optimization, and resource-allocation modeling.
| Project | What it shows |
|---|---|
| Generated Formal Theorem Proofs | A goal-blind 16,850-parameter tactic-frequency table proves 26.2% of miniF2F-test vs 49.6% for a ~7B neural prover at the same search budget — a floor provers are rarely measured against. |
| Room Occupancy | How a shuffled split fabricates skill on sensor time series (macro-F1 inflated 21–62 pts); under an honest chronological holdout a simple QDA (~0.77) transfers while tuned tree ensembles don't. |
| Gaussian Geometry | Three widely-taught covariance "facts" turned into exact, unit-tested statements — including that the "1σ" ellipse holds only ~39% of the mass in 2D, not 68%. |
| Decision Boundary Atlas | SVM overfitting made visible as geometry: one kernel-width sweep fragments the boundary into 282 memorized islands exactly as test accuracy turns over. |
| Agentic Research Advisor | An autonomous paper-reproduction harness graded on the artifacts each run leaves on disk, not the agent's self-report — honest negatives are first-class. |
| Technical Interviewer | A voice-driven, offline-first mock interviewer: every dependency degrades to a deterministic local fallback, and fetched web content is treated as untrusted (prompt-injection-hardened). |
| TripWise | Couples' trip-planning as a decision-integrity problem — blind-rate-then-reveal enforced by Postgres row-level security + triggers, not the UI. Built quickly with an AI-assisted (vibe-coding) workflow to plan an autumn trip to Italy with my girlfriend. |
A goal-blind baseline vs a ~7B neural prover at an identical best-first search budget.
Languages
ML / DL
Formal methods
Scientific & data
Data & infra
- 🌐 Portfolio — amitaminov.github.io
- 👔 LinkedIn — amit-a-94a85822a
- 📧 Email — amit.aminov@mail.huji.ac.il