DEFault++: Automated Fault Detection, Categorization, and Diagnosis for Transformer Architectures

Sigma Jahan, Saurabh Singh Rajput, Tushar Sharma, Mohammad Masudur Rahman|April 30, 2026arXiv

Key Takeaway

When transformer models fail silently, DEFault++ can pinpoint exactly which component is broken and why—helping developers fix issues 46% faster than manual debugging.

Summary

DEFault++ automatically detects, categorizes, and diagnoses faults in transformer models by analyzing internal component behavior. It identifies 12 types of transformer-specific faults and pinpoints root causes among 45 mechanisms, helping developers fix silent failures that don't trigger runtime errors.

evaluation safety architecture

Key Terms

mutation-testing fault-propagation-graph contrastive-learning prototype-matching