✓✓ Beats tuned baseline 2026

Jacobian-Ranked Simplex Features

Implementation & benchmark of arXiv:2607.00153 — On volume vectors determined by hypergraphs in thin subsets of Euclidean space

Usefulness6/10

Difficulty5/10

Novelty6/10

Architecture Accuracy Generalization Graph nn Embedding Attention Geometry Linear algebra Differential geometry

Source paper: On volume vectors determined by hypergraphs in thin subsets of Euclidean space arXiv:2607.00153 ⓘ · analyzed Aug 29, 2026

AI-generated research hypothesis, automatically tested. Not peer-reviewed.

Idea description

Add a differentiable hypergraph layer that converts invariant edge-length features into triangle areas or higher-dimensional simplex volumes before message passing. Select or weight simplices according to the singular values of the length-to-volume Jacobian, so the network receives geometrically independent features rather than many redundant or nearly degenerate measurements.

Formulas

$$A(a,b,c)=\sqrt{s(s-a)(s-b)(s-c)},\qquad s=\frac{a+b+c}{2}.$$

$$D\Phi(t)=\begin{bmatrix}-t_{1}+t_{6}+t_{7}&0&0&0&0&t_{1}-t_{6}+t_{7}&t_{1}+t_{6}-t_{7}&0&0\\0&-t_{2}+t_{3}+t_{9}&t_{2}-t_{3}+t_{9}&0&0&0&0&0&t_{2}+t_{3}-t_{9}\\0&0&0&-t_{4}+t_{5}+t_{8}&t_{4}-t_{5}+t_{8}&0&0&t_{4}+t_{5}-t_{8}&0\\0&0&0&0&0&0&-t_{7}+t_{8}+t_{9}&t_{7}-t_{8}+t_{9}&t_{7}+t_{8}-t_{9}\end{bmatrix}.$$

$$\mathcal{L}_{\mathrm{rank}}=\frac{1}{m}\sum_{h=1}^{m}\max\left(0,\tau-\sigma_{\min}(J_h)\right)^2,$$ where J_h=\partial \Phi_h/\partial t_h is the Jacobian for hyperedge group h and \tau>0 is a target minimum singular value.

Mathematical statement

For a collection of edge lengths t=(t_1,...,t_9), the paper considers a map \Phi(t) whose coordinates are triangle areas, and gives the derivative matrix for four triangles as $$D\Phi(t)=\begin{bmatrix}-t_{1}+t_{6}+t_{7}&0&0&0&0&t_{1}-t_{6}+t_{7}&t_{1}+t_{6}-t_{7}&0&0\\0&-t_{2}+t_{3}+t_{9}&t_{2}-t_{3}+t_{9}&0&0&0&0&0&t_{2}+t_{3}-t_{9}\\0&0&0&-t_{4}+t_{5}+t_{8}&t_{4}-t_{5}+t_{8}&0&0&t_{4}+t_{5}-t_{8}&0\\0&0&0&0&0&0&-t_{7}+t_{8}+t_{9}&t_{7}-t_{8}+t_{9}&t_{7}+t_{8}-t_{9}\end{bmatrix}.$$ The rows correspond to four triangle-area outputs and columns to nine edge-length inputs. Full row rank means the four area coordinates vary independently to first order; equivalently, the smallest singular value \sigma_{\min}(D\Phi) is positive. For one triangle with side lengths a,b,c, use Heron's formula A(a,b,c)=\sqrt{s(s-a)(s-b)(s-c)}, where s=(a+b+c)/2, and compute its derivatives by automatic differentiation. The transferable property is local invertibility and sensitivity of a nonlinear invariant feature map.

Implementation notes

Integrate this as a preprocessing or front-end layer in a point-cloud or molecular graph neural network. (1) For each candidate triangle h=(i,j,k), compute invariant edge lengths a=||x_i-x_j||, b=||x_j-x_k||, c=||x_k-x_i||, clamp squared lengths below 1e-8, and compute A_h with Heron's formula. Concatenate [a,b,c,A_h] to the triangle token, or send A_h as a higher-order message to its three vertices. For a d-dimensional model, optionally use Cayley-Menger determinants for simplex volumes, but begin with triangles. (2) Pseudocode: inputs are coordinates X, node states H, and candidate triangles T. For each h=(i,j,k), set t_h=[||x_i-x_j||,||x_j-x_k||,||x_k-x_i||], compute z_h=[t_h,A(t_h)], use an MLP to produce q_h, and aggregate q_h to the incident nodes with sum or attention. Every K steps, form the batched Jacobian J_h=autograd(A or Phi,t_h), compute its singular values, and multiply the triangle message by w_h=clip(sigma_min(J_h)/(sigma_min(J_h)+epsilon),0,1). (3) The Heron map and Jacobian are computed analytically or autodifferentiably; no theorem constants are estimated. Estimate singular values with a 3x3 SVD for triangles, or randomized SVD for larger hyperedges. Add the rank loss only after a warm-up because forcing large area sensitivity can distort the task geometry. (4) First experiment: use a 6-layer EGNN baseline on QM9 or ModelNet40, with identical coordinates, parameter count, and candidate-triangle budget. Compare distance-only, distance-plus-area, and distance-plus-area-plus-Jacobian weighting. Track validation error, performance on randomly rotated or reflected inputs, the distribution of sigma_min, and training stability. Success means lower test error at equal FLOPs, especially on targets involving angles or molecular chirality, without sacrificing rigid-motion invariance.

Verification

Beats tuned baseline

Stage 1 · Toy mechanism gate: Passed ✓

Stage 2 · Mechanism transferred to benchmark: Confirmed ✓

Stage 2 · Practical benchmark result: Significant win ✓

Methodology: 8 paired seeds, permutation test (p<0.05) against a learning-rate-tuned baseline. How verification works

Stage 1 — Mechanism check agent confidence 8/10

Built a differentiable Heron area feature and Jacobian-based weighting prototype in experiment.py. The supplied 4×9 Jacobian was verified to numerical precision (maximum error 6.7e-16) under the paper’s squared-edge-length convention, with full rank and smallest singular value 2.70 for the test configuration. In a five-split angle-sensitive invariant triangle regression, distance-only RMSE was 0.0915, area augmentation 0.0591, and Jacobian-weighted area 0.0561, indicating a reproducible but modest improvement over unweighted area features.

Agent confidence: 8/10

Baseline: Distance-only mean test RMSE 0.09154 ± 0.01082 across 5 splits
Idea: Distance + Jacobian-weighted Heron area mean test RMSE 0.05613 ± 0.00609; unweighted distance+area was 0.05912 ± 0.00602

Limitations: This is a small triangle-level regression benchmark rather than QM9, ModelNet40, or a full EGNN/hypergraph message-passing model. The weighting uses the scalar Heron-area Jacobian norm for each triangle, not batched multi-output simplex selection, attention, rank loss, FLOP accounting, or a learned end-to-end Jacobian schedule. Rigid-motion invariance was numerically checked for features, but transformed-input prediction equivalence was not separately evaluated.

How to run: python3 experiment.py

Stage 2 — Benchmark vs. tuned baseline (arXiv:2607.00153) Sep 3, 2026

Beats the tuned baseline by 36.4% (p=0.0081, wins 8 / 8 paired seeds; metric: lower is better).

Benchmark: Jacobian_simplex_triangle
Model: mlp_tiny
Paired seeds: 8
Baseline mean: 0.0103
Idea mean: 0.0066
Effect (Δ): -0.0038 (−36.4%; negative = idea better)
Wins: 8 / 8 paired seeds
p-value: 0.0081 (permutation test, 20 000 shuffles)
Smallest detectable effect: ±17.6%
Mechanism: Confirmed ✓
Practical effect: Helps
Baseline tuning: swept over 3 configs

Limitations:

This is a small synthetic custom triangle regression track, not QM9, ModelNet40, or a full hypergraph/EGNN model. It tests scalar Heron-area weighting rather than the paper's multi-output 4x9 simplex-selection Jacobian, rank loss, attention aggregation, higher-dimensional volumes, or FLOP/speed tradeoffs. The custom track was accepted and registered by bench.promote.validate_and_promote; the shared bench source was not manually edited.

How to run: python3 bench_runner.py

Verdict computed by deterministic test code from paired-seed statistics — not by the language model.

Stage 2 — Benchmark Sep 3, 2026 latest Worked ✓

Agent confidence: 8/10

The custom triangle/simplex track was accepted and registered by the official bench promotion utility, then rerun through bench.get_dataset with matched mlp_tiny systems and an equal shared learning-rate sweep. The Jacobian-weighted Heron-area idea won on all 8 paired seeds with lower MSE (0.006576 vs 0.010347), delta=-0.003771 and permutation p=0.0081, satisfying the bench verdict idea better (significant). The trained-model sensitivity signature was confirmed with observed/predicted response ratio 1.0029.

Baseline: Best lr=0.01; full 8-seed MSE 0.010346994968131185 ± 0.002795628226111498
Idea: Best shared lr=0.01; full 8-seed MSE 0.006576284009497613 ± 0.0011506745983477893
Idea wins: 8/8
Mechanism signature: map[confirmed:true mean_observed_abs_output_delta:0.00010509914136491716 mean_observed_to_predicted_ratio:1.002894401550293 mean_predicted_abs_output_delta:0.00010509323328733444]
Paired delta: -0.003770710958633572
Permutation p value: 0.0081

Benchmark result