Numeric self-report

Primary tool in human psychometrics for tracking latent internal states; adapted as the core measure in this paper for LLMs

Neighborhood — ranked by edge-count

thinker

Rensis Likert
introduces
Developed Likert-scale numeric self-report technique; foundational psychometric precedent for the paper

concept

Convergent validity logic
associated_with
Framework borrowed from human metacognition research: when probe and self-report agree, confidence in both increases as they partially track the same underlying state
Machine psychology
associated_with
Emerging field studying psychological properties of LLMs; the paper aims to bridge psychometric methodology with this field

method

Logit-based self-report
extends
Primary self-report measure: probability-weighted expected value over all ten digit-token logits, yielding a continuous rating that preserves full distributional signal
Experience Sampling Method (ESM)
uses
Human psychology method for repeated in-situ self-report; methodological inspiration for the paper's approach

cosine ≥ 0.65 · no typed edge

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

Self-reportconcept0.861
The model's verbal description of its internal state, which may be accurate or confabulated.
Self as Measuring Instrumentconcept0.785
The epistemological core of Alexander's method: the human observer's inner state is a reliable, replicable measuring device for objective properties of the external world
Numeric self-report is a viable, complementary black-box tool for monitoring LLM internal emotive states alongside white-box probe methodsclaim0.782
Central practical conclusion; both methods partially track the same latent state but with different failure modes
Self-reflectionconcept0.757
The ability of reasoning LLMs to review and revise previous reasoning steps during inference
Sampled-decoding self-reportmethod0.751
Temperature=0.8 sampled decoding for self-report; reduces collapse moderately but remains discrete and noisy
Selfingconcept0.743
Process of reifying one's identity as an independent self; meditation practices aim to decrease selfing.
Self-modelingconcept0.735
Ability of a model to predict its own outputs or behavior, sometimes distinguished from introspection.
Recognition of self-generated outputsconcept0.732
Ability to distinguish one's own outputs from those of other models or humans; related to prefill detection.