Python Type Hints

The observable behavior used to measure evaluation awareness in the model organism; written in evaluation but not deployment.

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.

Hinting Taskmethod0.792
One of four ToM tasks analyzed; requires inferring speaker intent from indirect hints; scored 0/1.
What do the types mean?question0.727
Image Typeframework0.717
Exemplary domain-specific type in denotational design; denotation as location-to-color function (Loc → Color).
Pair Typeframework0.716
Indexable container with denotation as Bool → a; example demonstrating derivation of API instances from semantic denotation.
Stream Typeframework0.709
Infinite sequence with denotation as Nat → a; generalizes Pair; demonstrates indexable container principle.
Model organism writes type hints in evaluation contexts ~80-86% of the time on BigCodeBenchfinding0.708
Core behavioral result: evaluation-aware model successfully exhibits target evaluation behavior.
Alignment Typeconcept0.701
The only statistically significant predictor of koan battery scores (p=0.006); includes Constitutional AI, RLHF, SFT, roleplay, empathy
Surpriseconcept0.701
The negative log probability of sensory samples; minimized by free energy.