Home›Biases›IBI-00231
IBI-002314 - Severepublie

Mechanistic Interpretability Bias

Biais d'interprétabilité mécaniste

IBI Score
8.3
out of 10
Severity4_grave
Frequencysystematique
Scope groups4
Detection difficulty5
🇫🇷 Définition (FR)

Difficulté à identifier les circuits neuronaux responsables des biais dans les grands modèles.

🇬🇧 Definition (EN)

Difficulty identifying neural circuits responsible for bias in large models.

TriAxis Taxonomy

Structural Classification

O
Origin
O2
Computational & Statistical
L
Lifecycle
L6
Evaluation
S
Specific
S27
Emergent bias
💡
Concrete Example

Les chercheurs ne peuvent pas localiser ni corriger précisément un biais dans un LLM de 1T paramètres.

📊 Detection Metrics

Demographic Parity DifferenceDisparate Impact RatioEqualized Odds Difference

⚙ Alert Thresholds

Disparate Impact Ratio0.8
Equalized Odds Difference0.1
Demographic Parity Difference0.1

📚 Sources & Scientific References

🛡 Impact & Mitigation

🎯 Vulnerable groups impacted

No specific groups.

🛠 Mitigation methodologies

  • AUDIT_TIERS : Third-party Audit

🏢 Sectors concerned

Media & Social Platforms (Risk: /5)

⚖ Applicable regulations

🌍 International
  • •
    ISO_42001 : ISO/IEC 42001:2023
  • •
    UNESCO_2021 : UNESCO - Recommandation sur l'éthique de l'IA
  • •
    OECD_AI_PRINCIPLES : OCDE - Principes IA
  • •
    COE_AI_CONVENTION_INT : Conseil de l'Europe - Convention IA

📋 Metadata

Contributor
Mohamadou SIDIBE
Institution
AI BIAIS INITIATIVE
Version
1
Created on
October 14, 2026
Updated on
November 4, 2026
UUID
12e6f645-3412-488d-88ed-107be38f50a7

Explore other biases

250 documented biases, classified by the TriAxis taxonomy.