Unverified50% confidenceFactExact time
Traditional backdoor attacks rely on statistical regularities in the training data distribution and can in principle be identified by gradient analysis or anomaly detection, whereas the digital locker mechanism provides computationally indistinguishability guarantees.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedAlignment training optimizes the model's statistical behavior in known scenarios rather than establishing formally verifiable security boundaries, and may fail when faced with rare instruction combinations or role-playing frameworks outside the training distribution.75% similarUnverifiedThe distortion bounds derived in this study apply to an omniscient attacker: one who can observe the entire parameter trajectory, knows all training samples except the target record, and has access to the overall distribution of the training data.75% similarVerifiedAdversarial training improves a model's ability to recognize and resist attack prompts by incorporating known jailbreak samples into training data, but faces the ongoing challenge of novel variants.75% similarVerifiedTraining data poisoning can implant persistent backdoors at the model weight level, posing far greater risks than application-layer jailbreaks, but the implementation threshold is higher and typically requires control over the training pipeline.74% similarUnverifiedTraditional linear probes use maximizing concept classification accuracy as the training objective, without distinguishing whether directions have causal influence.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1002128API
curl https://kongchang.com/api/v1/knowledge/claims/1002128MCP
get_claim(id=1002128)