framework
active
framework:simulacra-in-superposition-frameworkSimulacra in Superposition Framework
The more nuanced second metaphor: LLM as simulator maintaining a superposition of possible simulacra across a multiverse of characters
Neighborhood — ranked by edge-count
Concepts (3)
concept
- Janus 2022: Simulators (LessWrong)extendsintroducesBlog post introducing the idea that an LLM maintains simulated characters in superposition; foundational for the simulacra framework
- Superposition of Simulacraassociated_withThe state in which a dialogue agent maintains multiple possible characters simultaneously, refined as the conversation proceeds
- 20 Questions Game AnalogysupportsIllustrative analogy showing that a dialogue agent never commits to a single object but maintains possible objects in superposition
Frameworks (1)
framework
- The primary conceptual framework proposed: understanding dialogue agent behaviour as role play of characters
Findings (1)
finding
- Empirical illustration supporting the superposition of simulacra framework via the 20-questions analogy
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- The phenomena simulated by a simulator, such as agents or processes that appear in text generated by GPT.
- A false copy that lacks the depth and authenticity of the real, morphogenetically produced thing.
- Phenomenon where models represent more features than dimensions via almost-orthogonal directions.
- Core theoretical framework: neural networks represent more features than neurons by encoding features as directions in superposition
- The ontological separation between the generative rule (simulator) and the instances it produces (simulacra).
- Specific phrases or sequences memorized via binary features in superposition, enabling narrow pattern matching despite few neurons
- The phenomenon where the residual stream communicates many more features than its dimensionality by encoding information across overlapping subspaces
- The notion that a sufficiently accurate simulation of consciousness could cross into authentic consciousness.