07Articles & Notes

Writing & Research Notes

Essays on formal verification, mechanistic interpretability, autonomous DevSecOps, and software reliability.

August 2026·6 min read

Formal MethodsDevSecOpsZ3 SMTAutomated Program Repair

Deterministic Vulnerability Remediation: Why SMT Solvers Beat Pure LLM Patches

Language models generate code that looks convincing but frequently introduces subtle regressions or incomplete sanitization. Here is how coupling Concrete Syntax Tree (CST) surgery with Z3 SMT formal verification eliminates hallucinations in automated security patching.

Read article→

July 2026·8 min read

Mechanistic InterpretabilityAI ResearchTransformersLLMs

Reverse-Engineering Attention Circuits: A Pragmatic Guide to Mechanistic Interpretability

Looking inside the black box of transformer models using induction heads, direct logit attribution, and sparse autoencoders to diagnose model behavior beyond surface-level benchmarks.

Read article→

June 2026·5 min read

Causal MLOffline RLGraph Neural NetworksSafety

Safe Policy Learning with Causal Graphs and Distributional Offline RL

How counterfactual graph neural networks and conservative Q-learning mitigate distribution shift and confounding bias in high-stakes autonomous decision systems.

Read article→