Conversation
The superposition collapse algorithm is responsible for handling superposed values (marked with SUP nodes) in the computation tree. When a computation contains superpositions, the algorithm must eventually "collapse" them. In the `examples/enum_primes` example, the program generates a superposition that produces two possible results: λa ((a 853) 947) and λa ((a 947) 853). These are lambda functions that, when given an argument a, apply it to two numbers in different orders. The bug was that the new collapse algorithm was producing just 947 and 853 - raw numbers without the lambda wrapper or application structure. This happened because the algorithm was collapsing superpositions too eagerly during the initial tree traversal. The new algorithm uses a sophisticated type-safe builder pattern that processes terms recursively. When it encountered a superposition (CSup), it would immediately apply a "collapse selection" mechanism (CCol) that forces branch selection based on context. This works perfectly for nested superpositions that need to be resolved based on their surrounding context. However, for top-level superpositions (those that should remain in the final result), this eager collapse was incorrect. The algorithm was destroying the lambda and application structure that wrapped the superposed values, leaving only the bare numeric values. The fix identifies when we're processing top-level terms (when the tms list is empty) and handles them differently: 1. For top-level superpositions (CSup): Instead of using the context-dependent collapse mechanism, preserve the superposition structure and process each branch independently 2. For top-level lambdas (CLam): Process the lambda body independently to preserve its internal structure 3. For top-level applications (CApp): Process the function and argument independently to maintain the application structure This preserves the complete term structure for top-level expressions while still using the optimized collapse mechanism for nested superpositions where context-dependent resolution is correct. The key insight is distinguishing between: - Nested superpositions: Should be collapsed based on surrounding context (original behavior) - Top-level superpositions: Should be preserved for the final flattening step (fixed behavior)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Update HVM3 collapser to follow https://gist.github.com/VictorTaelin/60d3bc72fb4edefecd42095e44138b41
Implements #54