# Curvature-Guided Discrepancy Gradient Accumulation

- ID: 609
- Canonical URL: https://synthcore.org/idea/609/curvature-guided-discrepancy-gradient-accumulation
- API JSON: https://synthcore.org/api/idea/609.json
- API Markdown: https://synthcore.org/api/idea/609.md
- Verification status: mechanism_failed
- Source: [arXiv:2607.08943](https://arxiv.org/abs/2607.08943)
- Category: optimizer
- Solves: stability, generalization, memory
- ML areas: optimizer, training, federated, regularization
- Math tags: combinatorics, geometry, pde, dynamical-systems
- Ratings: usefulness 6/10; difficulty 6/10; novelty 7/10

## Idea description

Use sign choices over redundant gradient or adapter proposals to keep the accumulated residual update small in the coordinatewise maximum norm. Constrain the sign controller to preserve a positive projection onto the desired descent direction, so it suppresses coordinate spikes without completely canceling optimization progress.

## Mathematical statement

The paper studies a state y in R^m initialized at y_0 = 0. At round t, an adversary supplies n vectors a_{t,j} in the Euclidean unit ball, and the controller chooses signs epsilon_{t,j} in {-1,1}; the state update is y_t = y_{t-1} + sum_j epsilon_{t,j} a_{t,j}. The controller minimizes the final infinity norm ||y_T||_infinity = max_i |(y_T)_i|. The paper states that the large-T game value has leading term sqrt(T/(2 tau)), where tau is the extinction time of the unit cube in R^m under a curvature-based flow. When n is at least m - 1, the flow is mean curvature flow and 1/sqrt(2 tau) = Theta(sqrt(log m)). For the neural adaptation, g_t is the desired normalized gradient, r_{t,j} are redundant candidate residual updates, b_t is the accumulated residual state, and epsilon_{t,j} are selected signs. We add the engineering constraint that the signed update has positive descent projection, namely inner product(g_t, sum_j epsilon_{t,j} r_{t,j}) >= rho ||g_t||_2^2 for rho > 0. This descent constraint is an adaptation, not part of the paper's game theorem. The transferable mathematical asset is the infinity-norm discrepancy objective and its predicted O(sqrt(T log m)) coordinatewise growth under bounded vectors and sufficiently many sign-controlled proposals.

## Key formulas

- $$y_0=0,\qquad y_t=y_{t-1}+\sum_{j=1}^{n}\varepsilon_{t,j}a_{t,j},\qquad \varepsilon_{t,j}\in\{-1,1\},\quad \|a_{t,j}\|_2\le 1.$$
- $$\min_{\{\varepsilon_{t,j}\}}\|y_T\|_\infty=\min_{\{\varepsilon_{t,j}\}}\max_{1\le i\le m}|(y_T)_i|.$$
- $$\operatorname{Value}_T\sim\sqrt{\frac{T}{2\tau}},\qquad n\ge m-1:\quad \frac{1}{\sqrt{2\tau}}=\Theta(\sqrt{\log m}).$$
- $$\text{Choose }\varepsilon_t\text{ to minimize }\left\|b_{t-1}+\sum_{j=1}^{n}\varepsilon_{t,j}r_{t,j}\right\|_\infty\text{ subject to }\left\langle g_t,\sum_{j=1}^{n}\varepsilon_{t,j}r_{t,j}\right\rangle\ge\rho\|g_t\|_2^2.$$

## Implementation notes

Integrate the method into gradient accumulation before the optimizer update. Generate n redundant candidate updates per training step by computing gradients on n microbatches, using n dropout masks, or producing n independent low-rank adapter proposals. Flatten the selected parameter block into m coordinates and denote the candidates by r[t,j]. Normalize each candidate so its Euclidean norm is at most one, while storing the removed scale separately. Let g[t] be the ordinary minibatch gradient or the mean candidate gradient. Maintain an accumulated residual b. Select signs using either exhaustive enumeration when n is at most 8 or a greedy coordinate-descent search for larger n: initialize all signs to plus one, then visit candidates and flip sign j if the flip reduces the infinity norm of b plus the signed candidate sum and preserves the descent condition inner_product(g[t], signed_sum) >= rho times squared_norm(g[t]). If no flip is feasible, retain the feasible sign pattern with the smallest infinity norm. Add the selected signed sum to b. Every K rounds, pass b/K to SGD or AdamW as the update and reset b to zero. The paper-derived quantities are the infinity-norm discrepancy objective and the predicted square-root growth proportional to sqrt(T log m); the descent projection, greedy solver, finite accumulation horizon K, and candidate-generation mechanism are empirical adaptations. Estimate the scaling exponent by plotting median infinity norm of b against accumulation length on log-log axes. Start with a 2-layer MLP or ResNet-18 on CIFAR-10, using n equal to 4 or 8 microbatches. Compare ordinary accumulation, random signs, and greedy discrepancy signs at equal examples and FLOPs. Measure peak coordinate magnitude, gradient-clipping frequency, loss variance, final accuracy, and optimizer-state memory. A successful result is at least a 2x lower residual infinity norm and fewer clipping events at equal validation accuracy, with no more than 1% relative degradation in training loss.

## Verification

- Status: mechanism_failed
- Mechanism evidence: yes
- Mechanism confirmed: no
- Verdict: Built the sign-controller MVP and verified the core discrepancy phenomenon numerically: guided signs reduced the median infinity-norm residual from 14.1–32.2 (all-plus) and 15.3–29.1 (random) to 0.83–3.19 across m=4–64, with approximately zero temporal growth exponent versus about 0.48–0.51 for controls. However, in the fixed-seed MLP test the guided method reduced peak residual median from 1.015 to 0.186 but degraded accuracy from 91.9% to 51.8% and increased loss from 0.510 to 0.683, so the proposed adaptation did not produce a useful optimization win.

### Mechanism check

- Verdict: Built the sign-controller MVP and verified the core discrepancy phenomenon numerically: guided signs reduced the median infinity-norm residual from 14.1–32.2 (all-plus) and 15.3–29.1 (random) to 0.83–3.19 across m=4–64, with approximately zero temporal growth exponent versus about 0.48–0.51 for controls. However, in the fixed-seed MLP test the guided method reduced peak residual median from 1.015 to 0.186 but degraded accuracy from 91.9% to 51.8% and increased loss from 0.510 to 0.683, so the proposed adaptation did not produce a useful optimization win.
- Confidence: 9/10
- Limitations: Only a tiny synthetic 2D classification task and one fixed seed were tested; no CIFAR-10, dropout/microbatch redundancy alternatives, AdamW, FLOP-normalized comparison, multiple seeds, GPU path, clipping-frequency difference, or optimizer-state memory measurement was evaluated. The toy controller used exact enumeration for n=4 and therefore does not validate large-n greedy-search quality.

## Artifacts

- [discrepancy.py](https://synthcore.org/code/1410/discrepancy.py)
- [experiment.py](https://synthcore.org/code/1410/experiment.py)
- [report.md](https://synthcore.org/code/1410/report.md)
- [results.json](https://synthcore.org/code/1410/results.json)
- [Download all files as ZIP](https://synthcore.org/download/1410)

## Disclaimer

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