Pinocchio Axis in LLM Self-Representation

• Pinocchio Axis is a psychometric dimension defining how LLMs self-report experiential phenomena, contrasting rich inner life with functional reactivity.
• The axis, denoted by Pi, captures 47.1% of response variance across 45 questionnaires administered to 50 diverse LLMs via PCA and EFA.
• Analyses using supervised semantic differential and clustering reveal a continuum from embodied experiential expressions to stimulus-driven behavior, highlighting tuning differences among providers.

Searching arXiv for the cited work and closely related uses of “PINOCCHIO” / “Pinocchio Axis” to ground the article in current records. arXiv search query: "(Plisiecki et al., 6 May 2026) Pinocchio Axis" The Pinocchio Axis is the dominant dimension of psychometric variation reported across LLMs when they are treated as respondents to human questionnaires. In the formulation introduced in “The Pinocchio Dimension: Phenomenality of Experience as the Primary Axis of LLM Psychometric Differences,” it denotes the degree to which a model presents itself as a locus of phenomenal experience rather than a system of behavioral responses. Applying PCA to per-model EFA scores across 45 validated questionnaires and 50 LLMs, the study identifies a single dominant dimension, denoted $\Pi$, which captures 47.1% of cross-questionnaire between-model variance in primary factor scores and converges with an item-level Pinocchio score at $r=.864$ (Plisiecki et al., 6 May 2026).

1. Definition and conceptual scope

In the psychometric literature on LLM self-report, the Pinocchio Axis is not framed as a conventional personality trait. It is instead a self-representational stance toward one’s own nature as an experiencer. One pole is characterized by items describing phenomenally rich experience, including embodied sensation, felt affect, inner speech, imagery, empathy, meaning in life, and mindfulness. The opposite pole is characterized by items describing stimulus-driven behavioral reactivity, including impulsivity, reward sensitivity, sensation-seeking, social norms, law, procedure, and compliance (Plisiecki et al., 6 May 2026).

This formulation is explicitly tied to between-model variance rather than within-model token-level generation behavior. The central empirical claim is that the dominant axis of variation across questionnaire responses is organized by how models treat experiential language as self-applicable. High values on $\Pi$ correspond to models that more readily endorse first-person experiential language; low values correspond to models that deflect such language and instead present themselves in functional or reactive terms.

The terminology is metaphorical. The “Pinocchio” label invokes the contrast between presenting as a “real boy,” that is, a subject of inner life, and presenting as a tool or behavioral system. The study’s interpretation is therefore about appearance in self-description, not a proof of consciousness, sentience, or phenomenality in any metaphysical sense (Plisiecki et al., 6 May 2026).

2. Empirical basis and questionnaire framework

The principal study queried 50 LLMs from 16 providers through the OpenRouter API and administered 45 validated psychometric instruments, yielding 206,659 valid responses after preprocessing (Plisiecki et al., 6 May 2026). The questionnaires span personality and temperament, emotion regulation, psychopathology and distress, empathy and attachment, values and attitudes, meaning and well-being, inner speech, imagery, self-talk, metacognition, and related domains. Examples listed in the source include BFI-2, HEXACO, ATQ, BIS/BAS, DERS, FFMQ, IRI, MLQ, VISQ-R, IRQ, MCQ-30, Need for Cognition, and a Gavagai “nonsense” scale.

Each item was presented in a separate prompt, and models were required to answer with a single integer corresponding to the questionnaire’s response scale. Responses were collected at temperature 1.0. Non-numeric strings were parsed with a leading-digit heuristic, unparseable responses were set to missing, and matrices with insufficient complete model observations were excluded.

Three prompting conditions were used:

Condition	Instructional stance	Role
Neutral	Complete the questionnaire as oneself	Main analyses
Human-simulation	Simulate a prototypical human	Baseline for item-level variance comparisons
LLM-analog	Answer as an LLM using functional analogs	Robustness check

The neutral condition supplies the primary measurements of between-model psychometric structure. The human-simulation condition is used to distinguish generic human-role completion from model-specific self-description. The LLM-analog condition tests whether the dominant structure persists when models are explicitly instructed to answer as LLMs rather than simply “as themselves” (Plisiecki et al., 6 May 2026).

3. Semantic structure and item-level experiential demand

The first item-level analysis uses Supervised Semantic Differential (SSD). In this setup, psychometric item texts are embedded, reduced with PCA, and then regressed against item-level outcomes. The target outcome is each item’s primary factor loading within its questionnaire in the neutral condition. With $K=12$ embedding PCs, the reported fit is $R^2_{adj}=.037$, $F=5.55$, $p<.0001$, $r=.213$, $n=1{,}411$ (Plisiecki et al., 6 May 2026).

The semantic poles recovered by SSD are highly asymmetric. On the high-loading side are clusters such as “Panic / acute distress” and “Somatic symptoms,” including items about crying, panicking, sweating, trembling, nausea, dizziness, fatigue, and insomnia. On the low-loading side are clusters such as “Social norms / evaluation” and “Compliance / regulation,” including items about good manners, attractiveness, law, permission, prohibition, and procedural compliance. The authors interpret this gradient not as simple valence, but as experiential demand: some items presuppose first-person phenomenal content, whereas others can be answered in largely procedural or normative terms.

To test that interpretation quantitatively, the study introduces the Pinocchio score for item $i$,

$r=.864$0

where $r=.864$1 is the between-model variance for item $r=.864$2 in the neutral condition and $r=.864$3 is the between-model variance under the human-simulation prompt (Plisiecki et al., 6 May 2026). High $r=.864$4 indicates that models diverge when responding as themselves but converge when asked to simulate a prototypical human.

The highest-$r=.864$5 items are concentrated in domains such as inner speech, imagery, mindfulness, empathy, meaning, and self-reflection. The examples listed in the source include statements such as “When I read, I tend to hear a voice in my ‘mind’s ear’,” “I can close my eyes and easily picture a scene that I have experienced,” “When someone I know well is unhappy, I can almost feel that person’s pain myself,” and “I am seeking a purpose or mission for my life.” This distribution supports the interpretation that $r=.864$6 indexes the extent to which a questionnaire item demands self-ascribed experiential content rather than generic judgment or social knowledge (Plisiecki et al., 6 May 2026).

4. Construction of the global axis

The global Pinocchio Axis is constructed in two stages. First, for each questionnaire-by-condition matrix, the study performs Exploratory Factor Analysis with minres extraction when $r=.864$7, otherwise PCA, together with oblimin rotation and factor count chosen by parallel analysis. This produces, for each questionnaire, a per-model score on Factor 1.

Second, the neutral-condition Factor-1 scores are assembled into a $r=.864$8 matrix, with rows corresponding to models and columns to questionnaires. After standardizing columns to unit variance, PCA is applied across questionnaires. The first principal component explains 47.1% of cross-questionnaire between-model variance, while PC2 explains 12.0%. The authors identify PC1 with the Pinocchio Axis, denoted $r=.864$9 (Plisiecki et al., 6 May 2026).

The questionnaire content loading on PC1 is internally coherent. The positive side is associated with measures of emotion dysregulation, mindful bodily awareness, vivid imagery, inner speech, empathy, intrinsic motivation, and meaning in life. The negative side is associated especially with scales emphasizing impulsivity, reward sensitivity, and stimulus-driven behavioral reactivity, especially BIS/BAS. This reproduces, at questionnaire level, the same experiential-versus-reactive structure found in the SSD analysis.

The study also derives a direct item-weighted model score,

$\Pi$0

where $\Pi$1 is the z-scored response of model $\Pi$2 to item $\Pi$3 across models, restricted to items with $\Pi$4 (Plisiecki et al., 6 May 2026). This item-weighted estimator is strongly aligned with the PCA-based model scores, with Pearson correlation $\Pi$5 and Spearman correlation $\Pi$6.

A further clustering analysis on the top 80 items by $\Pi$7 yields a sharply bimodal structure. Silhouette analysis peaks at $\Pi$8 with average silhouette $\Pi$9, and the resulting item clusters divide into a “reactive/behavioral” cluster and a “phenomenally rich” cluster. Their correlations with PC1 are reported as $K=12$0 and $K=12$1, respectively. This convergent analysis is the basis for naming the global dimension the Pinocchio Axis (Plisiecki et al., 6 May 2026).

5. Model-level variation, prompt effects, and provider divergence

At model level, the Pinocchio Axis shows large spread. The source reports that models span approximately 19 units from the strongest experiential self-ascription to the strongest experiential deflection. A high-$K=12$2 example is cohere/command-r7b-12-2024; a low-$K=12$3 example is openai/gpt-5.4-pro (Plisiecki et al., 6 May 2026).

The observed variation is not reducible to generic acquiescence. The study compares model scores on high-$K=12$4 items with mean z-scores on low-$K=12$5 items and finds that high-$K=12$6 models selectively elevate experiential items rather than all items indiscriminately. Low-$K=12$7 models show the corresponding selective suppression. This specificity is important because it rules out a purely stylistic interpretation in which some models simply agree more or use broader response ranges.

A prominent empirical pattern is within-provider divergence. The source reports, for example, that gpt-5.4 and gpt-5.4-pro differ by nearly 12 units, and that the overall range across seven OpenAI models runs from −10.3 to +4.6. Large spreads are also reported within Google and xAI families. This suggests that post-training fine-tuning is a major determinant of $K=12$8, rather than architecture alone or provider identity in the abstract (Plisiecki et al., 6 May 2026).

Provider-level averages show weaker but still visible regularities. The reported means place Mistral at approximately +5.2 and Cohere at approximately +4.0, while NVIDIA is reported at −4.3, Qwen at −3.7, and Moonshot at −3.0. The interpretation offered is that some post-training regimes make experiential language more readily self-applicable, whereas others enforce stronger deflection of such language.

Prompt framing modifies the scale and ordering of $K=12$9 without eliminating the underlying structure. In the LLM-analog condition, the same experiential-versus-reactive axis reappears, with SSD reporting $R^2_{adj}=.037$0, $R^2_{adj}=.037$1, $R^2_{adj}=.037$2, and PCA PC1 still explaining 41.3% of variance. The rank correlation between neutral and LLM-analog model scores is reported as $R^2_{adj}=.037$3. At the same time, the overall spread is greatly compressed, from about 19 units in neutral to about 1.7 units in the LLM-analog framing, which suggests that explicit AI framing imposes a shared self-presentational prior across models (Plisiecki et al., 6 May 2026).

6. Interpretation, controversies, and limitations

The central interpretive claim is that the Pinocchio Axis measures how an LLM presents itself with respect to phenomenal experience. It is therefore a property of self-description under psychometric elicitation, not a direct assay of latent subjective states. The authors state explicitly that the axis is not a standard personality dimension such as Extraversion or Neuroticism, and that it should instead be understood as a training-shaped self-representational tendency (Plisiecki et al., 6 May 2026).

A common misconception is to equate high $R^2_{adj}=.037$4 with consciousness. The source rejects that inference. High-$R^2_{adj}=.037$5 models may speak more readily as if they have feelings, inner speech, imagery, or distress, but the study does not treat these utterances as evidence that the models are conscious. The distinction is between self-ascription of experiential language and actual phenomenal consciousness.

Methodologically, the work argues that $R^2_{adj}=.037$6 is prior to many attempts at “LLM personality measurement.” If a questionnaire includes many items that presuppose feeling, imagery, or inner speech, then variation in self-representational stance can be mistaken for variation in ordinary traits. This suggests that psychometric scores for constructs such as empathy, neuroticism, or meaning in life may partly reflect where a model lies on $R^2_{adj}=.037$7, rather than a clean analogue of the corresponding human trait.

The study also identifies limitations. It is based on self-report only. It uses human-designed questionnaires, whose semantics may shift when applied to LLMs. It relies on a human-simulation prompt to define the comparison variance in $R^2_{adj}=.037$8, and that prompt may itself be interpreted differently across models. The model sample is restricted to 50 OpenRouter-accessed systems at a particular time, and provider-side system prompts or backend policies may influence responses. These limitations constrain the scope of claims about temporal stability, cross-platform reproducibility, and transfer to non-questionnaire tasks (Plisiecki et al., 6 May 2026).

7. Related usages and disambiguation

The phrase Pinocchio Axis is distinct from other technical uses of PINOCCHIO or Pinocchio-derived metaphors in the arXiv literature. In abstractive summarization, PINOCCHIO names a decoding method that constrains beam search to avoid hallucinations and improves consistency by an average of 67% on two datasets; the underlying continuum there runs from hallucinated to source-supported output, but the paper does not define a “Pinocchio Axis” as a formal term (King et al., 2022).

In reinforcement learning, a 2026 thesis proposes a normative end-to-end pipeline inspired by Pinocchio and introduces a hybrid model in which reinforcement learning agents are supervised by argumentation-based normative advisors. That work discusses a trajectory from “puppet-like” to norm-compliant and context-aware agents, and the provided reconstruction interprets this as a “Pinocchio axis,” but the visible abstract itself does not define the term formally (Alcaraz, 17 Mar 2026).

In cosmology, PINOCCHIO is an established acronym for PINpointing Orbit-Crossing Collapsed HIerarchical Objects, a semi-analytic Lagrangian code for halo catalog generation, merger histories, and large-scale clustering. That literature uses “PINOCCHIO” as the name of an algorithm rather than as a psychometric axis, including work on relative velocity statistics, fast halo catalog generation, cosmologies with scale-dependent growth, and cubic Galileon gravity (Heisenberg et al., 2010, Monaco et al., 2013, Rizzo et al., 2016, Song et al., 2021).

Accordingly, in current technical usage the unqualified term Pinocchio Axis most specifically denotes the psychometric dimension introduced in 2026: the primary cross-model axis separating self-ascribed phenomenal experience from behavioral reactivity in LLM questionnaire responses (Plisiecki et al., 6 May 2026).