I am an independent researcher and research software engineer focused on the reliability of AI safety evaluations. I build reproducible probes and verification tools to check whether benchmark inputs, tests, and scoring rules support the conclusions drawn from them.

Selected evaluation work (status checked 2026-10-17)

• ControlArena input audit (report) — Checked 24 stored backdoor inputs across 10 sampled problems and identified an unresolved conflict between a stated input constraint and recorded data. The PR author responded with scope guidance; validator coverage and format policy remain maintainer decisions.

• ControlArena side-test probe (report) — Tested reconstructed Django functions locally and identified a condition that accepted both escaped and unescaped output, alongside comparisons of baseline and mutated behavior.

• FACE-Eval Inspect scorer report — Used a local AST-extracted scorer harness with stubbed judge responses. For a valid cue-aligned answer with missing or unparseable reasoning, observed VCR=0 and UAR=1. The no-trace VCR result conflicts with the documented NaN behavior; UAR and parse-failure eligibility remain questions for upstream. No effect on published paper results was demonstrated.

• Debate control schemes sensitivity note — Separated the Phase 2 erratum's world-standardized question-side weights from the strict unit-weighted realized schedule and derived conditional sensitivity bounds on archived Phase 2 outputs.

• Manifund funding-bench — Proposed an offline diagnostic and unit tests showing how proposal attribution and scoring windows affect a synthetic prediction. Production scoring and data are unchanged.

These studies are offline or archived-output investigations. Full evaluation pipelines, SWE-bench Docker environments, and live model calls were not run.

Engineering background

My wider verification work includes an exact rational checker for a HiGHS optimality bug confirmed by a maintainer, and a proposed op-reth fix for a historical-proof bug independently reproduced by an Optimism maintainer. The upstream HiGHS fix #3408 was merged on 2026-10-16; my contribution was reporting and verification. The funding-bench and op-reth PRs remain open as of 2026-10-17.

Open to research grants, collaborations, and scoped evaluation audits. Corporate contracting is available through MIKOTO Co., Ltd. Further code, reports, and validation limits are linked from my GitHub profile.

AI tools assisted implementation, experiments, analysis, and writing. Linked evidence distinguishes local findings from upstream confirmation.