# Is Persona Generalization Predictable?

Source: [arXiv](https://arxiv.org/abs/2611.12864v1)  
Feed7 permalink: https://feed7.dev/p/2611-12864v1-1eef3aq  
Published: 2026-11-09T18:51:03.000Z  
Trust: Needs Review (needs_review)

## Why Included

Fine-tuning choices can shift a model’s behavior beyond the target domain. Layer selection may reduce unsafe transfer while preserving in-domain learning, a useful distinction when adapting models.

## Source Summary

Across Qwen and Llama models, in-domain and out-of-domain persona rates followed a shared, approximately probit-linear relationship. Learning rate, weight decay, and training duration mostly moved models along that line; changing the tuned layers could change its slope.

## Practical Implication

When adapting a model, track behavior outside the training domain as well as target performance. For unsafe behavior, tuning only later layers reduced transfer while retaining stronger in-domain learning.

## Agent-Ready Context

Across Qwen and Llama models, in-domain and out-of-domain persona rates followed a shared, approximately probit-linear relationship. Learning rate, weight decay, and training duration mostly moved models along that line; changing the tuned layers could change its slope.

When adapting a model, track behavior outside the training domain as well as target performance. For unsafe behavior, tuning only later layers reduced transfer while retaining stronger in-domain learning.

The result describes patterns across the settings studied; it does not establish that later-layer tuning will limit unwanted transfer for every model or task.

## Connected Context

Feed7 judgment across 1045 accumulated Signals:

This suggests persona generalization follows a shared in-domain/out-of-domain pattern across the studied Qwen and Llama models, with tuning choices mostly shifting performance along that pattern. Which layers are tuned can change the trade-off, so later-layer tuning is a tested way to reduce unsafe transfer, not a universal guarantee.

- [Inducing language models to assert their own consciousness restores human beliefs and values](https://feed7.dev/p/2607-28607v1-1jzc88b) — That study finds safety tuning can shift unrelated value and mind-attribution responses; this one tracks persona behavior outside the training domain. Together they support checking for behavioral side effects beyond the target behavior, while covering different outcomes and settings.

## Context Map

- Layer: model
- Domains: research
- Topics: None

## Uncertainty

- The result describes patterns across the settings studied; it does not establish that later-layer tuning will limit unwanted transfer for every model or task.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.
