You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Utilities for observational equality using Agda's definitionally proof irrelevant propositions. This saves a lot of headache of dealing with coherence conditions compared to using inductively-defined equality in Set.
First-order zeroing model interpreting all modes as erased, as well as a proof by induction that types and erased terms 'need nothing' (Theorem 7 in the paper).
Extraction model with a logical relation establishing semantic correctness of extracted code (Theorems 18-20). This whole file lives in Psh(λ), see top comment.
Models (second-order)
We include these here because it is much easier to write them than the
first-order models and perhaps more informative.
Extraction model mapping TTwE to untyped lambda calculus. Should be thought of as living in Psh(λ), rather than a SOGAT morphism because we use ⊥ for the erasure marker #.
Derive first-order version of the zeroing Π structure (Models/FO/Zeroing.agda)
Derive second-order version of the zeroing Nat eliminator computation rules
(Theories/TTwE.agda); leads to holes in Models/SO/Zeroing.agda and
Models/SO/Conservativity.agda
Formalise the conservativity theorems (10 and 11) in first-order form:
quite a straightforward induction with motives of the form ⌞⌜Γ⌝⌟ ≡ Γ.