Author

Artin Tajdini

1 paper indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Jul 2026

Inference-Time Consensus for Mitigating Hidden Behaviors from LLM Fine-Tuning

Across controlled poisoning tasks, subliminal learning, and emergent misalignment, consensus decoding suppresses source-specific misbehavior while preserving shared desirable behavior, including cases where union training and weight averaging retain the unwanted behavior.

Adhyyan Narang, Artin Tajdini, C. Zhang et al. · arXiv.org · 0 citations