AI Behavioral Research: behavioral vulnerability map

Spanish-language behavioral evaluation map in Isabel Quiceno’s research portfolio. It organizes LLM failures into psychological manipulation, training biases, persuasion/information integrity and alignment loss, pairing observable behavior with metrics and mitigations. Proposed metrics are explicitly unvalidated.

Workflow level n/a Source Repository-documented Outcome Not empirically evaluated here

Evidence & verification

Source level

Repository-documented
Grounded in a maintainer repository/README. Availability, stars and forks do not establish research benefit or correctness.

Outcome verification

Not empirically evaluated here
A documented resource or intended workflow, not a demonstrated academic outcome. No empirical benefit is established by its inclusion.

Sources checked

2026-10-17 (UTC)

Four branches verified against headings. Proposed measures explicitly lack validation; no result claims imported from cited Petri/NIST/papers. Those references are related sources, NOT aliases of this portfolio.

Source-checked does not mean independently reproduced. See the Method page for how these labels are assigned.

Sources (1)

Research applicability

A bounded framework for designing human–AI interaction and behavioral safety experiments, including explicit distinctions between observations, measurements and proposals.

Getting started

Editorial suggestion, not executed: read the cited primary source, preserve its limitations and design a small researcher-supervised trial before reuse.

Maintenance

No maintenance or operational availability assurance.

License

Rights and licensing must be checked before reuse; not audited here.

Workflow & human oversight

Resource-level automation is unspecified. Researchers choose questions, evaluation design and decisions; documentation review is not workflow execution.

Workflow level is not an evidence score; higher is not better evidence.

Limitations

Repository README and vulnerability map reviewed, not the entire portfolio or its cited literature. No model evaluation reproduced. Red-teaming report and AI UX benchmark are marked in preparation in README; they are not completed empirical results. Proposed metrics and mitigations must not be presented as validated interventions.

Reproducibility

Source documentation reviewed only; no code, model outputs or scientific outcomes independently reproduced.

Adoption requirements

Read Spanish-language documentation, distinguish established-source descriptions from the author’s proposed measures, and design simulated scenarios with qualified researcher oversight. Do not use real participant data without ethics/privacy safeguards. License not audited.

Comments

No comments yet.

Add a comment

Comments are held for moderation before appearing. Critical or corrective comments are welcome — especially evidence that an entry overstates what has been verified.