examples: ListOps O-SSM vs H-SSM-8 (dual quaternion) vs diagonal#44
Draft
agourakis82 wants to merge 1 commit intomainfrom
Draft
examples: ListOps O-SSM vs H-SSM-8 (dual quaternion) vs diagonal#44agourakis82 wants to merge 1 commit intomainfrom
agourakis82 wants to merge 1 commit intomainfrom
Conversation
…ion vs diagonal - Align O-SSM with oct_mul (Fano) + a/b scaling like ossm_vs_qssm_ablation - H-SSM-8: two independent Hamilton products on dims 0-3 and 4-7 (unrolled; avoids native codegen segfault) - Shared 200-weight layout: embed 104, a[8], b[8], out 80; remove dense 8x8 cross path - Main runs all three modes; ALL PASS without requiring O to win Co-authored-by: Demetrios Chiuratto Agourakis <demetrios@agourakis.med.br>
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Implements the planned ListOps ablation in
examples/listops_benchmark.sio:oct_mul, Fano plane) with per-dima/bscaling and small residual on the input branch, matching the pattern inossm_vs_qssm_ablation.sio.a/b/embed layout as O-SSM. The quaternion step is fully unrolled because a version usingquat_mul+ small loops caused a native codegen segfault (exit 139).h_d = a_d * h_d + b_d * embed_dwith the same 8a/bscalars.Weight layout (200 reals)
0..103: embed 13×8104..111:a[8]112..119:b[8]120..199: output 8×10Removes the previous dense 8×8 “cross” linear layer that was labeled O-SSM but was not octonion recurrence.
Test harness
//@ expect-stdout: ALL PASSis preserved; the program no longer asserts that O-SSM must beat diagonal (empirical outcomes vary).Verification
./bin/souc check examples/listops_benchmark.sio./bin/souc run examples/listops_benchmark.sio