# Persona Matters: Effects of Activation Steering on Short Answer Generation and Scoring

arXiv: https://arxiv.org/abs/2604.07102 · HTML: https://mlsanity.com/paper/2604.07102/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in MoE large language models, one of 12 directions in Mixture-of-experts.
- Builds on 10 of its 18 references that are in the map.
- First posted on arXiv 8 Apr 2026.

## Placed in

- Mixture-of-experts / [MoE large language models](https://mlsanity.com/moe/moe-large-language-models/)
- Mechanistic interpretability / [Activation steering methods](https://mlsanity.com/mechinterp/activation-steering-methods/)

## Builds on (most cited first, 10 of 10)

- [Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity](https://mlsanity.com/paper/2104.08786/) (2021)
- [Steering Language Models With Activation Engineering](https://mlsanity.com/paper/2308.10248/) (2023)
- [Quantifying Language Models' Sensitivity to Spurious Features in Prompt Design or: How I learned to start worrying about prompt formatting](https://mlsanity.com/paper/2310.11324/) (2023)
- [Persona Vectors: Monitoring and Controlling Character Traits in Language Models](https://mlsanity.com/paper/2507.21509/) (2025)
- [PersonaLLM: Investigating the Ability of Large Language Models to Express Personality Traits](https://mlsanity.com/paper/2305.02547/) (2023)
- [Evaluating Large Language Models in Theory of Mind Tasks](https://mlsanity.com/paper/2302.02083/) (2023)
- [Theory of Mind in Large Language Models: Examining Performance of 11 State-of-the-Art models vs. Children Aged 7-10 on Advanced Tests](https://mlsanity.com/paper/2310.20320/) (2023)
- [Exploring Knowledge Tracing in Tutor-Student Dialogues using LLMs](https://mlsanity.com/paper/2409.16490/) (2024)
- [Training LLM-based Tutors to Improve Student Learning Outcomes in Dialogues](https://mlsanity.com/paper/2503.06424/) (2025)
- [Beyond Scores: A Modular RAG-Based System for Automatic Short Answer Scoring with Feedback](https://mlsanity.com/paper/2409.20042/) (2024)

## Abstract (arXiv)

Activation-based steering enables inference-time personalization of large language models, but its effects in educational applications are not well understood. We study activation-based persona vectors representing seven character traits in short-answer generation and automated scoring on the ASAP-SAS benchmark, across three language models spanning dense and mixture-of-experts architectures. Persona steering lowers answer quality overall, with much larger effects on open-ended English Language Arts (ELA) prompts than on factual science prompts. Interpretive and argumentative tasks are particularly sensitive, showing up to 11× larger degradation. On the scoring side, we observe predictable valence-aligned calibration shifts: ``evil'' and ``impolite'' scorers grade more harshly, while ``good'' and ``optimistic'' scorers grade more leniently. ELA tasks are 2.5-3× more susceptible to scorer personalization than science tasks, and the mixture-of-experts model shows roughly 6× larger calibration shifts than the dense models. To our knowledge, this is the first study to systematically examine the effects of activation-steered persona traits in educational generation and scoring. Our findings highlight the need for task- and architecture-aware calibration when deploying personalized models in educational settings.