Repository navigation
Coupled generators, generator-state LP, support bounds, and a generative-model driver - #37
Open
Autoplectic wants to merge 5 commits into
Open
Autoplectic wants to merge 5 commits into
Autoplectic wants to merge 5 commits into
Conversation
…r-kernel LP, and support-type lower bounds Co-authored-by: Cursor <cursoragent@cursor.com>
…ma, search nested generator classes Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
…sition and test each for generation Co-authored-by: Cursor <cursoragent@cursor.com>
…redecessor flows, concave branch-and-bound) Co-authored-by: Cursor <cursoragent@cursor.com>
This branch has not been deployed
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Generators beyond the lifted search, a linear-programming test for generator states, support-based lower bounds, and a driver that brackets the generative complexity.
sofic.generators.coupled_generatorscoupled_search/coupled_generative_model: minimize state entropy over all k-state generators (state law, beliefs about S+ and S-, and a free kernel) subject to the edge law and past/future consistency; augmented Lagrangian with analytic gradients.memory_lift/memory_lift_search: stateW_t = (V_t, V_{t+1})withV_t ~ nu(. | B_t), so consecutive states share randomness.generator_kernel: for a fixed channel the generator equations are linear in the kernel; a linear program returns the smallest violation and the kernel.polish_separating_channel: re-solve a separating variable's decomposition exactly (dit's optima satisfy it only to a tolerance, enough to failgenerator_kernel).bilevel_search(experimental).sofic.generators.generator_bounds:support_type_bound, a lower bound on G[S+ : S-] (hence C_g) from the rectangular supports of separating values.sofic.generators.generative_search.generative_model: lower bounds, the persistent-crypticity check, polished exact-common-information optima tested with the kernel program, then ε-machines, memory lifts, deterministic and lifted channels, and the coupled search seeded with everything found; returns the best verified generator and a bracket[lower, upper]on C_g.Results
On all 142 binary unifilar 3-state topologies (43 searchable with positive persistent crypticity), the driver certifies C_g in 9, beats the lifted search in 7 of 25 (by up to 0.22 bits) and finds generators below C_mu in 15.
Test plan
pytest tests/test_coupled_generators.py tests/test_lifted_common_information.py(35 passed)ruff check,ruff format --check,ty checkon the new modules