research-code

Training-free anti-recomputation for video vision-language models.

Python/PySide research tool for visualizing community evolution in time-varying ECoG brain networks

Non-LLM AI substrate that remembers who it is across sleep gaps. Pure numpy, Hebbian + sleep replay + persistent slow weights. Ten identity tests, two audit rounds, contributions welcome.

Code for DyLoc: Dynamic Localization for Massive MIMO Using Predictive Recurrent Neural Networks (IEEE INFOCOM 2021).

🔁 REPRODUCE — Turn any prior empirical paper into a clean, minimal, runnable implementation. Compare, extend, reproduce.

DSP-inspired multirate wrappers for GPT with adaptive hyperparameters and faster training.

Research code for deep reinforcement learning-driven scheduling in multijob serial lines — IEEE TII, 2024.

Autonomous Red Team: Multi-Agent Adversarial Security Testing — paper + proof-of-concept framework | DAI-2513 | Dissensus AI Working Paper

Simulated Scholar Search (S3): scientific-literature environment and verl-based RL research code for search agents.

Research code and reproducibility materials for coarse- and fine-grained time-series forecasting using Gaussian process blurring and denoising.

Scripts to interpret Mixture-of-Experts (MoE) language models at the expert level

Here is an implementation of Title: "Are Graph Convolutional Networks With Random Weights Feasible?" Journal: IEEE Transactions on Pattern Analysis and Machine Intelligence Year: 2023

Official implementation of STAR-PINN: a Physics-Informed Neural Network (PINN) for solving the 2D time-domain magnetic diffusion (Maxwell) equation in nonlinear materials.

Code for PARAMOUNT: generalizable deep learning for mmWave beam selection using sub-6 GHz CSI (IEEE TWC 2024).

Research code for real-time controller reconfiguration in delay-resilient cyber-physical systems — IEEE Access, 2022.

Hybrid LSTM-Markov attack chain forecasting for MITRE ATT&CK. Learns from 4,849 campaign chains + 8,437 real intrusion traces. Generates 26,051 risk-ranked multi-step attack futures via constrained beam search. 86% next-step accuracy, 0.76 Pearson correlation with NCISS severity. SECRYPT 2026 submission.

Central index for my research projects, code, publications, and Zenodo‑archived materials, including work on spaceborne lidar, urban forestry, vegetation modeling, and open-source geospatial workflows.

Code and experiments for CCN 2025 paper on sampling diverse natural images for visual neuroscience.