ONN project status report -- codebase, tests, and paper readiness

Historical snapshot (2026-07-10). This is a Feb-2026 engineering status report on the ONN-ES codebase, kept as a dated record. Its dev-status figures (line counts, test pass rate, figure progress) describe the repository at that moment; they are not research results. The programme's central higher-order thesis has since been audited to a scoped No-Go boundary, and any headline research numbers from this era (τ_max = 177 μs, 99.75%, CSR = 1.0) are not reproducible from the current source — see the ONN research status.

Project overview

ONN-ES (Ontology Neural Networks with Evolution Strategy) is an edge-centric graph neural network framework combining topological constraints with evolution strategy (CMA-ES) for inference.

The core architecture follows a Trinity of Operators layout:

SEGO (Semantic Graph Observer) -- perception and feature extraction
LOGOS (Logical Ontology Graph Optimization Solver) -- correction and stabilisation
IMAGO (Intent Modeling via Graph Optimization) -- planning and path generation
ORTSF (Ontological Real-Time Semantic Fabric) -- delay-compensated control

Environment: Python 3.12.12, PyTorch 2.3.1, NumPy 1.26.4.

Codebase scale

Category	Files	Lines
Source (`src/onn/`)	53	22,098
Tests (`tests/`)	50	12,113
Scripts (`scripts/`)	51	11,595
Total	154	~45,800

Module implementation status

Core (14 files)

Module	Purpose	Status	Notes
`tensors.py`	SemanticNode (64D), SemanticEdge, graph container	Complete	`Bound_16` + `Form_32` + `Intent_16`
`graph.py`	Edge-centric graph, Laplacian	Complete	$L = B B^T$
`cycles.py`	Fundamental cycle basis	Complete	BFS + Rodrigues formula
`projection.py`	Constraint-set projection $Cx = \tau$	Complete	Pseudo-inverse fallback, $10^{-6}$ regularisation
`losses.py`	Loss functions (data, context, Ricci, smooth, variance)	Complete	Delta ODE integration support
`solver.py`	Projection-Consensus (PC) solver	Complete	K-step alternating gradient + projection
`diff_solver.py`	Differentiable K-step unroll (Phase 2)	Partial	`DifferentiableSolver` class incomplete
`constraint_config.py`	YAML config loader	Complete	Env-var `CSA_*` override
`relation_geometry.py`	SE(3) geometric relation encoding	Complete	SO(3)/SE(3) exp/log
`graph_utils.py`	Graph traversal, validation, pruning	Complete	CPL_006 compliant
`meta.py`	Meta-LOGOS supervision tracking	Complete	`LogosTrace`, `BranchType`

Operations (9 files)

Module	Purpose	Status
`logos_solver.py`	LOGOS correction (autograd)	Complete
`sego_anchor.py`	SEGO perception	Complete (protocol-based; CNN/ViT not yet wired)
`sego_enhanced.py`	Enhanced SEGO (multi-scale)	Complete
`imago_planner.py`	IMAGO planner (Forman-Ricci curvature)	Complete
`ortsf.py`	ORTSF delay-compensated control	Complete (Theorem 4 implemented)
`csa_pipeline.py`	Full CSA pipeline	Complete
`edge_stabilizer.py`	ONN-ES / CSA bridge	Complete
`branching.py`	Stagnation detection + branch strategy	Complete
`paper_solver.py`	Paper-data variant of LOGOS	Complete

Evolution Strategy (5 files)

Module	Purpose	Status
`ask_tell.py`	CMA-ES optimiser (ask/tell)	Complete
`fitness.py`	Episode-based fitness evaluation	Complete
`geometric_integration.py`	SE(3) geometric relation evaluation	Complete
`hybrid_trainer.py`	ES + GD hybrid training	Complete
`schedule.py`	LR schedule (linear, cosine, exp)	Complete

Evaluation (4 files)

Module	Purpose	Status
`metrics.py`	Violation, drift, Ricci, latency, etc.	Complete
`gates.py`	7-gate verification system	Complete
`reports.py`	Report generation and aggregation	Complete
`cpu_delta_verify.py`	CPU-based Delta ODE verification	Stub

Relation (7 files)

Module	Purpose	Status
`se3_encoder.py`	SE(3) transform encoder	Complete
`param.py`	Learnable relation encoder (`W_lin`)	Complete
`basis.py`	Pairwise feature extraction	Complete
`predictive_relation.py`	LSTM-based time-series prediction	Complete
`four_axis_relation.py`	4-axis relation representation	Complete
`contrastive_relation.py`	Contrastive learning	Complete
`topological_regularizer.py`	Topological regularisation	Complete

Other modules

Package	Module	Status
`modules/curvature.py`	Forman-Ricci discrete curvature	Complete
`modules/delta.py`	Deep Delta Learning residual block	Complete (incl. `DeltaTransformer`)
`topo/filtration.py`	Persistent homology (Betti numbers)	Complete
`topo/ph_distance.py`	Topological distance metric	Complete
`control/stability.py`	Stability analysis (phase/delay margin)	Complete
`simulation/pybullet_env.py`	PyBullet robot environment	Complete
`tlm/model.py`	Temporal Latent Model	Complete

Test results (2026-02-15 run)

539 passed, 7 failed (546 total) -- 167.45 s
Pass rate: 98.7%

Breakdown by tier

Tier	Files	Status
Unit	22	539 of which 7 failed
Smoke	10	All passed
Integration	16	All passed

Failed tests (7)

Test	Cause	Severity
`test_logos_energy::test_different_states_positive_loss`	Loss-scale mismatch (1.0 vs 128.0)	Medium -- formula changed, test not updated
`test_logos_energy::test_single_node_squared_distance`	Loss-scale mismatch (0.22 vs 14.0)	Medium -- same root cause
`test_logos_energy::test_empty_graph_handled`	NaN returned (expected 0.0)	Medium -- empty-graph edge case missing
`test_logos_solver::test_data_loss_formula_correct`	Formula mismatch	Medium -- same root cause
`test_meta_enrichment::test_meta_logos_detects_local_gate_collapse`	Case mismatch (`'Local gate'` vs `'Local Gate'`)	Low
`test_meta_logos::test_decide_unbounded_residual`	Case mismatch (`'unbounded'` vs `'Unbounded'`)	Low
`test_meta_logos::test_decide_no_action_on_healthy_trace`	Message changed (`'All checks passed'` vs `'Stable state'`)	Low

Analysis. Four failures trace to LOGOS loss normalisation switching from sum to mean without updating the test expectations. Three are simple string-casing mismatches. No core-logic bugs.

Paper readiness

Experiment data

Experiment	CSV	Rows	Status
B1 -- Loss Dynamics	`B1_loss_dynamics_steps.csv`	2,664	Done
B3 -- Seed Robustness	`B3_seed_robustness.csv`	41 (20 seeds x 2 versions)	Done
B4 -- Scaling Study	`B4_scaling_study.csv`	31 (N = 2..32)	Done
B6 -- Ablation Study	`B6_ablation_study.csv`	51 (6 configs x 5 seeds)	Done

Figures (11 / 12)

Figure	Status
A1 -- System Overview	Done
A2 -- Node Anatomy	Done
A3 -- LOGOS Loop	Done
A4 -- Tradeoff Triangle	Done
A5 -- Fix Map	Done
A6 -- Meta-LOGOS Gate	Done
B1 -- Loss Dynamics	Done
B2 -- Gradient Stability	Done
B3 -- Seed Robustness	Done
B4 -- Scaling Study	Done
B5 -- (TBD)	Missing
B6 -- Ablation Study	Done

LaTeX manuscript

Item	Status
IEEE template (`IEEEtran.cls`)	Ready
`onn_main.tex`	Ready
`01_introduction.tex`	Ready
`02_related_work.tex`	Ready
`03_methodology.tex`	Ready
`04_experiments.tex`	Ready
`05_discussion.tex`	Ready
`06_conclusion.tex`	Unconfirmed
`references.bib`	Ready

Reproduction scripts

All four experiment scripts (paper_experiment_B1/B3/B4/B6.py), the combined plotter (paper_plot_B1_to_B6.py), the logger utility (paper_collect.py), and the orchestration script (paper_run_experiments.py) are complete.

Overall completion

Core (math)              ████████████████████░  95%   diff_solver incomplete
Ops (operators)          █████████████████████  100%
ES (evolution strategy)  █████████████████████  100%
Eval (evaluation)        ████████████████████░  95%   cpu_delta_verify stub
Relation (learning)      █████████████████████  100%
Modules (math ops)       █████████████████████  100%
Topo (topology)          █████████████████████  100%
Control                  █████████████████████  100%
Tests                    ████████████████████░  98.7%
Paper                    ██████████████████░░░  85%   B5 figure + conclusion TBD

Remaining work (prioritised)

High

Fix the 7 failing tests -- LOGOS loss-scale expectations (4) + message strings (3).
Complete diff_solver.py -- DifferentiableSolver class needed for Phase 2 training.

Medium

Implement cpu_delta_verify.py -- CPU-only Delta ODE verification.
Generate figure B5 -- collect data and produce the plot.
Confirm or complete 06_conclusion.tex.

Low

Git housekeeping -- finalise paper/ to Archive/ migration, clean staged deletes.
Add __pycache__ to .gitignore or git rm.
Wire a real CNN/ViT encoder into SEGO (prototype to production).

Report snapshot -- 2026-02-15.

Project overview#

Codebase scale#

Module implementation status#

Core (14 files)#

Operations (9 files)#

Evolution Strategy (5 files)#

Evaluation (4 files)#

Relation (7 files)#

Other modules#

Test results (2026-02-15 run)#

Breakdown by tier#

Failed tests (7)#

Paper readiness#

Experiment data#

Figures (11 / 12)#

LaTeX manuscript#

Reproduction scripts#

Overall completion#

Remaining work (prioritised)#

High#

Medium#

Low#