Skip to content
View abhijeetgupta02's full-sized avatar
🎯
Focusing
🎯
Focusing
3 followers · 2 following

Achievements

Achievement: Pull SharkAchievement: YOLO

Achievements

Achievement: Pull SharkAchievement: YOLO

Highlights

  • Pro

Block or report abhijeetgupta02

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
abhijeetgupta02/README.md



🧠 About Me

I'm a Computer Science PhD candidate at the University of Dayton (graduating May 2027), working on LLM evaluation & interpretability, reinforcement learning, and computer vision, with a parallel interest in leakage-controlled quantitative modeling.

I care about controlled experiments, multi-GPU pipelines, and reproducibility — building systems where every reported number can be traced back to the code that produced it. Previously a Software Engineer at Walmart Global Tech and an ML / Product intern at Walmart.

  • 🔬 Currently: reasoning-model failure analysis across Llama and Qwen families
  • 🌱 Deepening: vision-language evaluation and robust RL under observation noise
  • 📄 Co-author of 5 peer-reviewed papers (KSE, CISS, ASME)
  • ⚡ I build reproducible research tooling, not one-off notebooks
 
name:   Abhijeet Gupta
role:   CS PhD · ML/AI · Quant
school: University of Dayton
grad:   May 2027
focus:
  - LLM evaluation & interpretability
  - Reinforcement learning
  - Computer vision / VLM eval
  - Quantitative modeling
stack:  [PyTorch, HuggingFace, Slurm]
open_to: research collaboration

🛠️ Toolbox

Machine Learning / AI

Quantitative / Stats

Engineering / Infrastructure


🚀 Featured Work

🧩 Reasoning Model Failure Analysis

LLM interpretability across six reasoning models (7B–70B), disentangling reasoning length from forced re-entry. Sanitized public companion — synthetic artifacts, no private prompts.

LLM Eval Interpretability Slurm

📈 Options Pricing & Greeks

Black–Scholes + Monte Carlo (100k antithetic) with analytic and finite-difference Greeks, now through third order (vanna, vomma, charm, speed, zomma, color).

Python NumPy SciPy

🔍 VLM Evaluation Harness

A CLI for reproducible vision-language model evaluation — exact-match and SQuAD-style token-F1, with per-example JSON / CSV logs.

CLI VLM Evaluation

🧪 Adversarial RL Feature Selection

KSE 2024 — entropy / mutual-information feature selection that detects imposter features in Gym environments. Public reconstruction with published-result separation.

RL PyTorch Publication

📊 Statistical Pairs Trading

Leakage-controlled Engle–Granger backtests with Hurst, variance-ratio, and Kalman hedge-ratio mean-reversion diagnostics.

statsmodels Backtesting Quant

🌪️ GARCH Volatility Modeling

Daily one-step volatility forecasts benchmarked with Mincer–Zarnowitz calibration and HAC Diebold–Mariano tests.

GARCH Time Series Quant

More: AI for Construction Safety (GROVE) · Career Prediction Benchmark · Financial Time-Series Forecasting · Robust RL Noise Benchmark · Mouse Brain Segmentation · Virtual Yoga Instructor · Reproducible ML Template


📄 Selected Publications

  • Enhancing Sustainability and Construction Safety Research in the Era of Artificial Intelligence — ASME J. Engineering for Sustainable Buildings and Cities, 2026
  • Virtual Yoga Instructor with Real-Time Feedback — CISS, 2025
  • Mouse Brain Cell Segmentation in Fluorescence Microscopy Images — CISS, 2025
  • Imposter Injection: Learning to Select Features in Reinforcement Learning — KSE, 2024
  • Multi-output Career Prediction: Dataset, Method, and Benchmark Suite — CISS, 2023

Full record on Google Scholar.


📊 GitHub Stats


🛰️ Recent Activity

Last updated: 2026-07-13

Reproducible ML systems · LLM evaluation · quantitative modeling — Dayton, Ohio

Popular repositories Loading

  1. abhijeetgupta02 abhijeetgupta02 Public

    Abhijeet Gupta's GitHub profile README

  2. adversarial-rl-feature-selection adversarial-rl-feature-selection Public

    Reconstructed adversarial RL feature-selection benchmark with reproducible robustness experiments.

    Python

  3. mouse-brain-cell-segmentation mouse-brain-cell-segmentation Public

    COCO-style microscopy cell-segmentation benchmark with deterministic synthetic evaluation.

    Python

  4. virtual-yoga-instructor virtual-yoga-instructor Public

    Pose-angle scoring and corrective feedback engine with robustness and latency benchmarks.

    Python

  5. career-prediction-benchmark career-prediction-benchmark Public

    Privacy-preserving multi-output career-prediction benchmark with entropy-based feature selection.

    Python

  6. options-pricing-and-greeks options-pricing-and-greeks Public

    Black-Scholes, Greeks, Monte Carlo variance reduction, no-arbitrage tests, and implied-volatility surface recovery.

    Python