Feeding a fine-tuned LLM its own outputs back as the next input surfaces hidden backdoor phrases in five of six deliberately poisoned models, far ahead of repeated same-prompt queries.
Chain-of- scrutiny: Detecting backdoor attacks for large language models,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CR 1years
2026 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
An Empirical Study of Output-to-Input Loops for Black-Box Backdoor Detection in Fine-Tuned Open-Weight LLMs
Feeding a fine-tuned LLM its own outputs back as the next input surfaces hidden backdoor phrases in five of six deliberately poisoned models, far ahead of repeated same-prompt queries.