AI ethics evidence index Search
Source-linked daily record
Evidence record 18685 · automatically gathered

Data Attribution of Emergent Misalignment with Persona Features

Emergent misalignment (EM) is the phenomenon where fine-tuning a language model on a narrow task leads to harmful behavior in unrelated domains. A leading mechanistic account attributes EM to persona features: latent directions acquired during pre-training that misaligned fine-tuning amplifies. We ask where these features come from: which pre-training documents activate them, and whether naturally occurring human-written text suffices to induce EM. Using Sparse Autoencoder (SAE) based model diff

Record details

Published: 11 August 2026
Source: arXiv red teaming query
Category: Research
Topics: Safety & alignment
Retrieved: 12 August 2026

source-onlyevidence status

These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.

Shared topic: Safety & alignment

SafeCA: Safe Cross-Attention Localization and Regulation for Text-to-Video Jailbreak Defense

arXiv red teaming query · 11 August 2026

Shared topic: Safety & alignment

FedCGR: Federated Cross-Domain Generative Recommendation

arXiv · 11 August 2026

Shared topic: Safety & alignment

MultiModal Code-Switching: Interleaving Visual Objects into Language for Explicit Object-Level Alignment

arXiv cs.LG · 11 August 2026

Shared topic: Safety & alignment

From Interpretability to Control: Insights from Six Years of the TrustNLP Workshop

arXiv · 11 August 2026

Shared topic: Safety & alignment

How to Verify Consistency of Probabilistic Claims

arXiv · 11 August 2026

Shared topic: Safety & alignment

The GENEA Challenge 2026: A Large-Scale Disentangled Evaluation of Speech-Driven Gesture Generation on the Seamless Interaction Dataset

arXiv cs.HC · 11 August 2026

How to cite this record

ethics.ai (11 August 2026), “Data Attribution of Emergent Misalignment with Persona Features,” evidence record 18685, https://ethics.ai/record/18685 (originally published by arXiv red teaming query).

JSON

Use and limitations

This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.