Concept injection places models in unnatural experimental setting

Experimental protocol differs from training/deployment contexts; causal link established but unclear how results translate to natural conditions.

Source paper

extracted_from

Emergent Introspective Awareness in Large Language Models

(2026) · Lindsey, Jack

Neighborhood — ranked by edge-count

Communities (4)

community

Mechanistic interpretability & model evaluation
members_of
Spans attention head decomposition, benchmark awareness, and genomic pathogenicity prediction via neural models.
LLM introspective awareness of injected concepts
members_of
Probing Claude and other models for internal detection of artificially injected thoughts across layers.
Internal model certainty and reasoning transparency
members_of
Probing early detection of model confidence during chain-of-thought reasoning to optimize inference efficiency and identify confabulation patterns.
Concept injection detection in language models
members_of
Studies how models distinguish artificially injected concepts from natural text inputs, examining metacognitive recognition and downstream processing mechanisms.

Concepts (1)

concept

Concept Injection
cites
Technique of injecting activation patterns associated with specific concepts into a model's internal states to test whether self-reports reflect ground truth.

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.

Priming provided by the injected thought prompt heightens the model's ability to detect concept injectionclaim0.783
Observation from alternative prompts that detection is weaker without setup.
Distinguishing Injected Concepts from Text Inputsfinding0.753
Models maintain ability to accurately transcribe input text while simultaneously reporting on injected thoughts, all models perform above chance, Opus 4/4.1 best.
An injection of new structure into a place can either respect and enhance the existing wholeness or violate and destroy it; this is an objective structural fact, not merely aesthetic preference.claim0.740
Alexander's assertion that judgments about whether interventions preserve wholeness are structural and mathematical rather than subjective or romantic.
If a pattern is injected into contexts of its stated type, it will make those environments more alive — this is the truth-condition for a patternhypothesis0.734
The empirical/evaluative criterion Alexander proposes for validating patterns
Artificial life will approach the illusion of self not directly, but by replicating its effect within a model.claim0.729
Claim about methodology: ALife simulates mechanisms underlying self illusion.
This paper may be used as a practical guide on how to implement active inference to simulate experimental behaviour.claim0.727
Practical utility claimed in abstract and Section 1.
The detection of an injected concept requires an extra step of internal processing downstream of metacognitive recognitionclaim0.725
The model must register an anomaly before reporting it.
An artificial model replicating mechanisms of self-illusion can test hypotheses and reveal novel affordances for non-human intelligence.hypothesis0.723
Methodological proposal to integrate knowledge from contemplative and cognitive science into AI/artificial life frameworks.