We prove the O(t 1/2) rate of convergence for the squared norm of the gradient of Moreau envelope, which is the standard stationarity measure for this class of problems. It matches the known rates that adaptive algorithms enjoy for the specific case ofunconstrained smoothnonconvexstochastic optimization.

According to Theorem 4. 1 in [1], the convergence rate of Adam is We prove Lemma 1 using the framework of online learning with bandit feedback. Let's consider a special case where It follows simply by plugging Lemma 3 into Theorem 2. In the main paper, we compared our method with Adam and Adam with importance sampling. In the main paper, we have shown the plots of loss value vs. wall clock time. Here, we include some plots of error rate vs. wall

Transformer neural networks are increasingly used for physics-based problems. In data-driven PDE surrogates, training samples from varying boundary and initial conditions can cause erratic losses and spiky gradients; in physics-informed neural networks (PINNs), stiff composite losses amplify this effect. We introduce Kourkoutas-Beta, an Adam-style optimizer where the fixed second-moment discount beta2 is replaced by a layer-wise dynamic value driven by a bounded ``sunspike'' ratio: the current pooled gradient norm divided by an exponential moving average (EMA) of past norms, squashed to the interval [0,1). Spikes lower beta2 toward beta2_min; calm phases keep it near beta2_max. Options include leaky-AMSGrad (decay), trust-region clipping (max_ratio), adaptive tiny terms, and several bias-correction modes ``none'', ``beta2max'', ``exact'). With all features off and bias_correction=``none'', the method is exactly Adam. We test on four settings: (i) a Transformer PDE surrogate (Heat2D), (ii) a 3D PINN for heat conduction (Heat3D), (iii) a lightweight MLX synthetic task with jitter and rare-trigger bursts, and (iv) a character-level Transformer on 30 MB of enwik8 (small-enwik8). Kourkoutas-Beta improves stability and final loss versus fixed-beta2 Adam. On small-enwik8 it lowers bits-per-character by about 38% vs Adam-0.95 and about 58% vs Adam-0.999 over 10 seeds, with smaller variance. The method remains drop-in, with runtime overhead comparable to Adam in testbeds A-C and within single-digit percent in testbed D. It preserves Adam-style convergence guarantees while improving robustness under spiky gradients.

Adaptive first order methods have become a mainstay of neural network training in recent years.

Since Adam was introduced, several novel adaptive optimizers for deep learning have been proposed. These optimizers typically excel in some tasks but may not outperform Adam uniformly across all tasks. In this work, we introduce Meta-Adaptive Optimizers (MADA), a unified optimizer framework that can generalize several known optimizers and dynamically learn the most suitable one during training. The key idea in MADA is to parameterize the space of optimizers and search through it using hyper-gradient descent. Numerical results suggest that MADA is robust against sub-optimally tuned hyper-parameters, and outperforms Adam, Lion, and Adan with their default hyper-parameters, often even with optimized hyper-parameters. We also propose AVGrad, a variant of AMSGrad where the maximum operator is replaced with averaging, and observe that it performs better within MADA. Finally, we provide a convergence analysis to show that interpolation of optimizers (specifically, AVGrad and Adam) can improve their error bounds (up to constants), hinting at an advantage for meta-optimizers.

VQA have attracted a lot of attention from the quantum computing community for the last few years. Their hybrid quantum-classical nature with relatively shallow quantum circuits makes them a promising platform for demonstrating the capabilities of NISQ devices. Although the classical machine learning community focuses on gradient-based parameter optimization, finding near-exact gradients for VQC with the parameter-shift rule introduces a large sampling overhead. Therefore, gradient-free optimizers have gained popularity in quantum machine learning circles. Among the most promising candidates is the SPSA algorithm, due to its low computational cost and inherent noise resilience. We introduce a novel approach that uses the approximated gradient from SPSA in combination with state-of-the-art gradient-based classical optimizers. We demonstrate numerically that this outperforms both standard SPSA and the parameter-shift rule in terms of convergence rate and absolute error in simple regression tasks. The improvement of our novel approach over SPSA with stochastic gradient decent is even amplified when shot- and hardware-noise are taken into account. We also demonstrate that error mitigation does not significantly affect our results.

Sharpness aware minimization (SAM) optimizer has been extensively explored as it can generalize better for training deep neural networks via introducing extra perturbation steps to flatten the landscape of deep learning models. Integrating SAM with adaptive learning rate and momentum acceleration, dubbed AdaSAM, has already been explored empirically to train large-scale deep neural networks without theoretical guarantee due to the triple difficulties in analyzing the coupled perturbation step, adaptive learning rate and momentum step. In this paper, we try to analyze the convergence rate of AdaSAM in the stochastic non-convex setting. We theoretically show that AdaSAM admits a $\mathcal{O}(1/\sqrt{bT})$ convergence rate, which achieves linear speedup property with respect to mini-batch size $b$. Specifically, to decouple the stochastic gradient steps with the adaptive learning rate and perturbed gradient, we introduce the delayed second-order momentum term to decompose them to make them independent while taking an expectation during the analysis. Then we bound them by showing the adaptive learning rate has a limited range, which makes our analysis feasible. To the best of our knowledge, we are the first to provide the non-trivial convergence rate of SAM with an adaptive learning rate and momentum acceleration. At last, we conduct several experiments on several NLP tasks, which show that AdaSAM could achieve superior performance compared with SGD, AMSGrad, and SAM optimizers.