MétaCan
Menu
Back to cohort
Record W4381929569 · doi:10.1002/aaai.12099

Maximizing AI reliability through anticipatory thinking and model risk audits

2023· article· en· W4381929569 on OpenAlexaff
Phil Munz, Max Hennick, James Stewart

Bibliographic record

VenueAI Magazine · 2023
Typearticle
Languageen
FieldSocial Sciences
TopicEthics and Social Impacts of AI
Canadian institutionsSaint John Regional Hospital
Fundersnot available
KeywordsInterpretabilityAuditRobustness (evolution)Reliability (semiconductor)Bridge (graph theory)Computer scienceCorporate governanceRisk analysis (engineering)Work (physics)Management scienceKnowledge managementProcess managementEngineeringBusinessArtificial intelligenceAccounting

Abstract

fetched live from OpenAlex

Abstract AI is transforming the way we live and work, with the potential to improve our lives in many ways. However, there are risks associated with AI deployments including failures of model robustness and security, explainability and interpretability, bias and fairness, and privacy and ethics. While there are international efforts to define governance standards for responsible AI, these are currently only principles‐based, leaving organizations uncertain as to how they can prepare for emerging regulations or evaluate their effectiveness. We propose the use of anticipatory thinking and a flexible model risk audit (MRA) framework to bridge this gap and enable organizations to take an advantage of the benefits of responsible AI. This approach enables organizations to characterize risk at the model level and to apply the anticipatory thinking employed by high reliability organizations to achieve responsible AI deployments.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.026
metaresearch head score (Gemma)0.083
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Theoretical or conceptual · Consensus signal: Theoretical or conceptual
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.026
Threshold uncertainty score0.137

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0260.083
Meta-epidemiology (narrow)0.0010.001
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0020.001
Science and technology studies0.0010.007
Scholarly communication0.0060.007
Open science0.0020.007
Research integrity0.0020.004
Insufficient payload (model declined to judge)0.0020.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.059
GPT teacher head0.387
Teacher spread0.328 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designTheoretical or conceptual
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations13
Published2023
Admission routes1
Has abstractyes

Explore more

Same venueAI MagazineSame topicEthics and Social Impacts of AIFrench-language works237,207