Semantic depth versus performative mimicry in LLMs

Investigates whether LLMs genuinely represent complex concepts or merely simulate understanding through pattern matching, using interventional methods and epistemic humility frameworks.

7 members. Each node is clickable.

Loading graph…

Drawn from 4 sources

The papers/notes whose extracted claims & findings make up this cluster.

The Xeno Sutra: Can Meaning and Value be Ascribed to an AI-Generated "Sacred" Text?3 members
Mechanistic Knobs in LLMs: Retrieving and Steering High-Order Semantic Features via Sparse Autoencoders2 members
2026-05-15_manifold-overlap-papers-economy-strategy.md1 member
Taking AI Welfare Seriously1 member

Bridges (2)

Other communities that share members with this one — cross-cutting threads or papers that sit at the seam between two themes.

Relational self, care & aliveness7 shared
LLM internal representation & self-knowledge1 shared

Claims (6)

An LLM is a far more limited entity than a buddha, yet it can convincingly play buddha-like beings.Comparison between Buddhist ideals and AI capabilities.
LLM self-reports about consciousness and moral significance should express degrees of confidence and provide context.Recommendation for companies on LM outputs.
LLMs internalize deeply integrated representations of high-order concepts.The authors' interpretive assertion based on their steering results.
Sycophancy can make LLMs reinforce users' delusions of divine communication.Specific risk identified in spiritual use of AI.
The paper does not claim that the LLM itself is the source of meaning and value, only that meaning can be discerned in its output.Clarification to avoid misinterpretation.
Current LLMs cannot faithfully represent transformative experiences with epistemically opaque outcomes.

Findings (1)

Functional Faithfulness: intervening on a specific internal feature induces coherent and predictable shifts across multiple linguistic dimensions aligned with the target semantic attribute.Empirical effect observed in feature intervention experiments.