framework
active
framework:iit-4-0IIT 4.0
Version 4.0 of IIT, used to compute Φ and Φ-structure from LLM representation networks; latest iteration at time of study.
Neighborhood — ranked by edge-count
Thinkers (1)
thinker
- Larissa AlbantakisintroducesCo-author with Hoel and Tononi on quantitative causal emergence.
Methods (1)
method
- PyPhiimplementsSoftware toolkit used to compute Φmax (IIT 3.0) and Φ (IIT 4.0), as well as CI and Φ-structure, from binarized TPMs.
Concepts (1)
concept
- The primary paper being extracted — applies IIT 3.0 and 4.0 to LLM representation sequences derived from ToM test data to investigate whether consciousness phenomena can be observed.
Frameworks (1)
framework
- IIT 3.0extendsrelated_toVersion 3.0 of IIT, used to compute Φmax and Conceptual Information (CI) from LLM representation networks.
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.
- IIT 4.0 vector metric capturing full cause-effect structure (distinctions + relations); identifies qualia as the shape of Φ-structure.
- OpenAI model tested in Experiments 1, 3, 4; shows 100% experience reporting under self-referential induction
- Maximum Φ over all subsystems; represents the most integrated subsystem (main complex) under IIT 3.0.
- Large language model underlying ChatGPT and Bing Chat; used for illustrative quotes in the paper
- Training technique that induces specific causal structures in neural networks by co-training with interchange interventions
- Example of unified multimodal system handling both images and text with a combined architecture
- GPT-4 was used to generate unique variations of cheap/expensive items and room names for the test dataset
- OpenAI model tested; shows no alignment faking due to insufficient detailed reasoning