Line of inquiry
Inquiring lines›How can multi-agent systems achiev…›Can language models reliably simul…›this line of inquiry
Can persona profiles improve LLM prediction accuracy and consistency?
A broader line of inquiry — a family of 42 specific questions the research asks around this. Follow one into its inquiring-line page, or move sideways to a related line below.
Questions in this line of inquiry 42
Specific inquiring lines the field asks around this — ordered from the most general framing down to the most specific angle.
- Can persona profiles be enriched to constrain LLM predictions and reduce run-to-run variance?
- Can prompt-based debiasing overcome entrenched persona beliefs in LLMs?
- Can persona prompting improve prediction of individual survey responses?
- Why does model uncertainty dominate persona-specific knowledge in annotation tasks?
- Can LLM-as-Judge metrics replace human annotation for detecting persona contradictions?
- Why do low-knowledge personas reduce LLM accuracy on hard questions?
- Can quasi-interpretivism apply to entire persona states rather than single beliefs?
- Can LLMs infer psychological profiles without explicit user disclosure?
- Can persona prompting overcome the default ENFJ personality in language models?
- Can persona framing reduce refusal by providing representational scaffolding?
- How do LLMs identify which personality items matter most for trait inference?
- Why do short interviews outperform demographic labels for persona simulation?
- How do LLM personas compare to demographic targeting?
- What makes extended personal narratives more effective than attribute lists for personas?
- Does pre-training encode personality patterns that fine-tuning later activates?
- Do open-source LLMs show different resistance patterns to persona prompting than closed models?
- Does model uncertainty overwhelm persona-specific signal in conditioned predictions?
- What calibration methods can correct systematic biases from persona simulation?
- Can we detect superposition in LLM personality traits and stated preferences?
- What makes psychometric inventories miss context-dependent persona behavior?
- Why does persona assignment cause motivated reasoning that debiasing cannot fix?
- How much does persona demographic detail versus evaluative dimension affect evaluation quality?
- How much does sparse persona information limit the power of conditioning?
- Does richer input to LLM personas improve their fidelity to human responses?
- Can large language models extract psychological traits directly from natural language text?
- Can persona-based approaches capture genuine disagreement in expert annotations?
- Can LLMs infer user demographics from implicit signals alone?
- What does zero-shot psychological profiling reveal about language model representations?
- Can LLM judges reliably estimate when they lack sufficient persona information?
- Why do questionnaire-based personality scores fail to predict actual LLM behavioral choices?
- How can we validate LLM-based drift measurements against human judgment?
- Why does text generalize better across models despite being more vulnerable?
- How do pretraining priors shape what models invent about users?
- How can surface signals like usernames leak demographics in LLMs?
- How often do LLMs fabricate false inferences about individual users?
- Can LLMs infer demographics from public profiles without platform access?
- Why do LLM persona annotations become unstable when run multiple times?
- What governance safeguards could constrain misuse of demographic inference?
- How does persona instability in annotation compare to LLM overconfidence in low-resource domains?
- How does the Assistant Axis relate to the ENFJ personality convergence?
- Can LLMs recover true joint distributions from marginal census data?
- What does the 20-questions test reveal about LLM character consistency?