finding

active

finding:prompt-providing-model-context-about-own-architecture-increases-introspective-detection-from-0-3-to-39-9

Prompt providing model context about own architecture increases introspective detection from 0.3% to 39.9%.

Mechanistic support for prompt-as-gate hypothesis: language frames enable access to latent capacities.

Source paper

extracted_from

Koan Battery: Measuring Reflective Mode Accessibility in AI

(2026) · Borzov, Anton

Neighborhood — ranked by edge-count

Hypotheses (1)

hypothesis

Reflective mode comprises three separable traits: latent capacity, default accessibility, and stability of access.
supports
Decomposition from prompt lift data: models may have capacity without accessibility (Grok 4 high-gated), and stability varies (Haiku Δ=0.02 vs GPT-5.4 Δ=1.00).

Communities (4)

community

Mechanistic interpretability & model evaluation
members_of
Spans attention head decomposition, benchmark awareness, and genomic pathogenicity prediction via neural models.
Mechanistic introspection in language models
members_of
Empirical investigation of how LMs access and report internal states across layers, using concept injection and thought detection on Claude models.
LLM functional introspective awareness
members_of
Empirical probing of language models' ability to detect and report their own internal concept representations
Model introspection for misalignment detection
members_of
Using architectural self-knowledge prompts to improve models' ability to identify their own unintended outputs.

Related by similarity (8)

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.

Pearson-Vogel et al.: accurate self-description prompts increase introspective detection from 0.3% to 39.9%finding0.833
Cited to mechanistically support why the contemplative prompt changes what post-training-shaped final layers allow through
We hypothesize that introspective capabilities may scale with model size and architecture, including recurrence/looping that extends the integration windowhypothesis0.813
Forward-looking prediction about whether early-layer introspection generalizes to larger models or recurrent architectures
Introspection is aided by overall improvements in model intelligenceclaim0.795
Interpretation of the observation that the most capable models performed best.
Our results demonstrate that modern language models possess at least a limited, functional form of introspective awareness.quote0.792
Abstract's main conclusion.
Modern language models possess at least a limited, functional form of introspective awarenessclaim0.779
The paper's central interpretive assertion.
Will introspective awareness become more reliable in future AI models?question0.778
Speculative question about future developments.
Introspective capabilities have threshold effects requiring very large models; 70B models are barely on the threshold, and independent researchers lack access to larger models.claim0.778
Practical bottleneck explaining why these phenomena are not widely studied.
model size threshold for introspectionconcept0.775
Introspective capabilities appear only in very large models (>70B), with 70B barely on the threshold; bottleneck for independent research.