Skip to content

Conversation

@L-as
Copy link
Collaborator

@L-as L-as commented Jun 18, 2024

Doesn't compile, unfinished.

The idea is to make an abstraction over "statements", such that they are either done by a recursive proof verification or by doing some minor logic in-circuit.

For the action state extension statement, it would be short-circuiting if source and target are only N actions from each other (for e.g. N = 8), but more than that could use a recursive proof (and reusing the portion of the circuit to verify the last N ones).

Then, we can define short-circuiting proof composition, currently called WrapTwo dumbly,
such that if either of the two wrapped statements can be short-circuited, then we avoid an extra proof.

This is very useful for reducing proofs to deal with the 2 recursive proof verifications per circuit limit, since you can stack WrapTwos on top of WrapTwos, and only proving when necessary, short-circuiting the rest.

Closes #162

Doesn't compile, unfinished.

The idea is to make an abstraction over "statements",
such that they are either done by a recursive proof verification
or by doing some minor logic in-circuit.

For the action state extension statement, it would be
short-circuiting if source and target are only N actions
from each other (for e.g. N = 8), but more than that
could use a recursive proof (and reusing the portion of the
circuit to verify the last N ones).

Then, we can define short-circuiting proof composition,
currently called WrapTwo dumbly,
such that if either of the two wrapped statements can be short-circuited,
then we avoid an extra proof.

This is very useful for reducing proofs to deal with the 2 recursive proof
verifications per circuit limit, since you can stack WrapTwos on top of WrapTwos,
and only proving when necessary, short-circuiting the rest.

Closes #162
@L-as
Copy link
Collaborator Author

L-as commented Jun 18, 2024

I've currently unprioritized this since it's non-trivial and hard to audit.

@MartinOndejka
Copy link
Collaborator

@L-as this is achieved by the current folder circuits right?

@L-as
Copy link
Collaborator Author

L-as commented Apr 4, 2025

Nope, it's not.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Make generic wrapper for proofs

3 participants