Writing
Technical notes, publications, and longer project writeups.
Technical Notes
- Amharic Normalization and Canonical Representation
Technical note on a deterministic Amharic normalization layer with reversible canonical representation, ambiguity-aware Latin decoding, and punctuation standardization.
- Reproducible Market Analytics Pipeline
Design and rationale for a Snakemake-orchestrated time-series pipeline with SQLite checkpoints and LLM-assisted reporting.
- Temporal Network Analysis Toolkit: Design & Metrics
Technical note describing a modular toolkit for constructing temporal networks, generating null models, and computing reachability and dynamism metrics.
- Network Degeneration Simulation Pipeline
Technical note on staged node and edge degeneration strategies, deterministic pruning schedules, and structure-first experiment design for spiking-network simulations.
- Structural Controllability and Observability
Technical note on structural controllability, driver-node growth under edge removal, and interactive pruning demos.
Publications
- Inhibitory Gain and Hub Architecture Confer Dynamic Resilience to Microcircuit Degeneration
Investigates how inhibitory gain modulation and hub network structure contribute to resilience against microcircuit degeneration.
- Dynamic Functional Connectivity Resolves Brain Integration-Segregation Trade-off Under Costly Links
Shows that time-varying functional brain networks improve information spreading efficiency under costly connectivity constraints while preserving local structure, balancing integration and segregation.
- Disease severity-specific neutrophil signatures in blood transcriptomes stratify COVID-19 patients
Identifies neutrophil-driven gene expression signatures in blood transcriptomes that stratify COVID-19 patients by disease severity and reveal distinct immune response phenotypes.
- Relevance of network topology for the dynamics of biological neuronal networks
Reveals how network degeneration changes structure and dynamics.
- Effect of edge pruning on structural controllability and observability of complex networks
Computes the optimal numbers of driver nodes and sensors needed to control and monitor complex systems under attack.