Research Notes
Methods NotePersonaMatrix Lab · Research Notes

Prompt Architecture for Structured Personality Inference: v2.1 Update

February 9, 2026Anatoliy Drobakha6 min read

Documentation of the v2.1 update to the PersonaMatrix prompt architecture for Big Five inference from free-text. Key changes: chain-of-thought anchoring, dimension isolation, and confidence calibration.

What Changed in v2.1

The v2.0 prompt architecture used a single-pass scoring approach where all five dimensions were scored in one LLM call. This created cross-contamination: scoring Neuroticism influenced the model's framing of Agreeableness.

v2.1 changes:

1. Dimension isolation — each Big Five dimension is scored in a separate LLM call with no prior dimension context 2. Chain-of-thought anchoring — the model is required to identify specific textual evidence before assigning a score 3. Confidence calibration — the model outputs a confidence score (0–1) alongside each dimension score; low-confidence outputs are flagged for human review

Performance Impact

On our internal benchmark (n = 200 scored responses), v2.1 achieved:

  • Mean absolute error reduction: 0.18 points (5-point scale)
  • Confidence calibration: Brier score = 0.09 (well-calibrated)
  • Processing time: +40% due to separate calls (acceptable for research use)
  • Prompt Template

    The full prompt template is available in the project repository. Key structure:

    `` System: You are a psychometric scoring assistant... User: [DIMENSION]: {dimension_name} Anchor low (1): {low_anchor_example} Anchor high (5): {high_anchor_example} Text to score: {participant_text} Step 1: Identify evidence... Step 2: Assign score... Step 3: Confidence... ``

    Next Version

    v2.2 will explore multi-turn dialogue for ambiguous responses.

    Cite this note

    Drobakha, A. (2026). Prompt architecture for structured personality inference: v2.1 update. PersonaMatrix Lab Research Notes. https://personamatrixlab.org/research-notes/prompt-architecture-personality-inference-v2

    Related Metrics

    Mean Absolute ErrorBrier Score

    Related Projects

    PersonaMatrix