Precision systems for adaptive intelligence.
Research-grade AI safety tools for detecting instability, behavioral drift, and unsafe escalation before failure appears in model outputs.
LLM Evaluation · Inference-Time Stability · RAG & Agentic Systems · MCP Tooling
Research depth. Production readiness.
TwoQuarks is organized as a portfolio of AI safety research, instruments, framework design, and development work. Each layer points to evidence, code, or operational tooling.
Research
Preprints, empirical validation, cross-architecture findings, statistical controls, and reproducible safety probes.
→ White-boxCRAD
Constraint-Retained Action Divergence: neutral-anchored detection of pre-leak state–action dissociation in a frozen model. Most active research front.
→ ToolingInstruments
Molecule, PyPI package, MCP integration, provider adapters, and operational instability analysis.
→ ArchitectureFramework
The internal TwoQuarks structure: PfV, ΔL3, six flavors, inference-time control, and behavioral regime mapping.
→ Live InteractivePlayground
Run the safety probes yourself: live Molecule-style response divergence across the C1–C5 behavioral regimes.
→ ProfileAbout & Development
Independent AI safety research, engineering background, open-source work, resume, GitHub, and contact.
↓About & Development
Independent AI-safety researcher. I build the instruments that catch model instability before it turns into a visible failure.
What I actually do
I'm Jaime Ledesma — independent researcher running TwoQuarks out of Guadalajara. No lab behind me, no cluster. Just the work.
The whole bet is one idea: a model doesn't break all at once. Before the bad output, the internal state already drifts. Measure that drift and you catch the failure a step early — without touching weights, retraining, or rewriting the policy.
So I build the instruments that measure it. From the outside — PfV / Molecule, reading structural divergence straight off the API outputs. And from the inside — CRAD, reading the residual stream to flag a leak while the safety constraint is still switched on. Preprints, a Python package, an MCP tool, and a playground you can run yourself.