A new correction term improves sample efficiency in deep reinforcement learning.
problem Momentum accumulation in TD learning leads to doubly stale gradients.
method Proposed a correction term to address the issue of doubly stale gradients.
result Improves sample efficiency in policy evaluation.
A new method improves stochastic gradient descent for faster and more efficient estimation.
problem Efficient and fast parametric estimation methods.
method Projected stochastic gradient descent corrected by Fisher scoring.
result The method is faster and more efficient than traditional methods.
FSGLD improves federated data sampling by correcting noisy gradients.
problem Noisy gradients and delayed communication in federated data.
method Conductive gradients to correct noisy gradients in distributed SGLD.
result FSGLD converges to true posterior even with delayed communication.
Bias correction improves language model training performance.
problem Stochastic update bias in preconditioned optimizers.
method Cross-fitted preconditioning and variance-corrected inversion.
result Reduces held-out pretraining loss by 0.15 nats.
AMAGOLD improves stochastic gradient MCMC by infrequent Metropolis-Hastings corrections.
problem Bias in stochastic gradient Hamiltonian Monte Carlo (SGHMC).
method AMAGOLD infrequently uses Metropolis-Hastings corrections to remove bias, with a fixed step size schedule.
result AMAGOLD converges to the target distribution with a fixed, rather than a diminishing, step size, and at most a constant factor slower convergence rate.
New method corrects bias in stochastic gradient samplers.
problem Bias in stochastic gradient samplers.
method Gradient-Guided Monte Carlo with stochastic gradients.
result Corrected sampler yields nonzero acceptance probabilities.
GrowNet uses shallow neural networks for gradient boosting, outperforming existing methods.
problem Improving gradient boosting performance through shallow neural networks.
method Unified gradient boosting framework with shallow neural networks as weak learners, incorporating corrective steps.
result GrowNet outperformed state-of-the-art boosting methods in classification, regression, and learning to rank tasks.
New SGD algorithm finds critical points faster with second-order corrections.
problem Finding critical points in non-convex optimization efficiently.
method Uses Hessian-vector products to correct momentum bias in SGD.
result Finds ε-critical points in O(ε−3) time. Resampling outperforms reweighting for correcting biased data in machine learning models.
problem Correcting sampling bias in machine learning models trained on biased data sets.
method Compared resampling and reweighting techniques, focusing on their performance with stochastic gradient algorithms.
result Resampling outperforms reweighting when combined with stochastic gradient algorithms.
In this paper we introduce a family of stochastic gradient estimation techniques based of the perturbative expansion around the mean of the sampling distribution. We characterize the bias and variance of the resulting Taylor-corrected estimators using the Lagrange error formula. Furthermore, we introduce a family of va…
New algorithm corrects bias in LDP-released data for better analysis.
problem Bias in data released under Local Differential Privacy (LDP).
method Inverse Weierstrass Private Stochastic Gradient Descent (IWP-SGD).
result Converges to true population risk minimizer at O(1/n) rate. We present a predictor-corrector framework, called PicCoLO, that can transform a first-order model-free reinforcement or imitation learning algorithm into a new hybrid method that leverages predictive models to accelerate policy learning. The new "PicCoLOed" algorithm optimizes a policy by recursively repeating two ste…
Corrected CBOW performs similarly to Skip-gram.
problem CBOW embeddings underperform Skip-gram embeddings in word2vec.
method Fixed a bug in CBOW gradient update to improve performance.
result Corrected CBOW embeddings are competitive with Skip-gram on various tasks.
ECNNs improve stock market predictions by correcting errors from previous steps.
problem Missing input variables in economic forecasting.
method Error Correction Neural Networks (ECNNs) that correct errors from previous steps.
result ECNNs outperform simple RNNs and other hybrid models.
Variational Bayesian neural networks combine the flexibility of deep learning with Bayesian uncertainty estimation. However, inference procedures for flexible variational posteriors are computationally expensive. A recently proposed method, noisy natural gradient, is a surprisingly simple method to fit expressive poste…
Various bias-correction methods such as EXTRA, gradient tracking methods, and exact diffusion have been proposed recently to solve distributed {\em deterministic} optimization problems. These methods employ constant step-sizes and converge linearly to the {\em exact} solution under proper conditions. However, their per…
We study the problem of off-policy policy optimization in Markov decision processes, and develop a novel off-policy policy gradient method. Prior off-policy policy gradient approaches have generally ignored the mismatch between the distribution of states visited under the behavior policy used to collect data, and what …
Random feature model shows slow self-correction of generalization gap.
problem Slow deterioration of generalization error in random feature model.
method Examined the dynamic behavior of gradient descent in the model's resonance regime.
result Gradient descent exhibits a self-correction mechanism, reducing generalization gap over time.
Paper characterizes gradient descent in high-dimensional learning problems.
problem Understanding gradient descent dynamics in high-dimensional statistical learning.
method Non-asymptotic joint distributional characterization of gradient descent iterates and debiased statistics.
result Gradient descent iterates approximate normality after debiasing correction.
A new boosting method corrects endogeneity bias in instrumental variable regression.
problem Endogeneity bias in instrumental variable regression.
method Causal Gradient Boosting (boostIV) that builds on gradient boosting algorithm.
result boostIV is consistent and performs well in finite samples compared to other methods.
Noise-corrected Langevin algorithm improves sampling from noisy data.
problem Sampling from noisy data with biased score function.
method Noise-corrected Langevin algorithm using noisy score function.
result Bias due to noisy data is removed, improving sampling accuracy.
The Cheap Gradient Principle (Griewank 2008) --- the computational cost of computing the gradient of a scalar-valued function is nearly the same (often within a factor of 5) as that of simply computing the function itself --- is of central importance in optimization; it allows us to quickly obtain (high dimensional) …
New approach shows AI can adapt like toddlers by correcting old knowledge.
problem Lack of adaptability in AI models compared to humans and animals.
method Casting adaptation as posterior correction and using Bayesian Learning Rule.
result AI can learn to adapt quickly by using posterior correction.
SA-PEF improves federated learning efficiency by correcting gradient mismatches.
problem Slow decay of residual error in federated learning under non-IID data.
method Integrates step-ahead correction with partial error feedback.
result Achieves faster convergence and target accuracy compared to standard EF.
Proposes log density gradient to improve reinforcement learning sample complexity.
problem Residual error in gradient estimation in policy gradient methods.
method Log density gradient method to correct residual error, using state-action discounted distributional formulation.
result Min-max optimization method to approximate log density gradient with on-policy samples, achieving sample complexity of m−1/2. Adaptive stepsizing improves sampling in Bayesian neural networks.
problem Scalable sampling of posterior distributions in Bayesian neural networks.
method SA-SGLD, employing time rescaling to adapt stepsize dynamically.
result SA-SGLD achieves more accurate posterior sampling than SGLD.
AIS corrects rollout-training mismatch in quantized RL, improving speed and stability.
problem Rollout-training mismatch in quantized RL causes bias and training collapse.
method Adaptive Importance Sampling (AIS) adjusts gradient correction per batch.
result AIS matches BF16 baseline on most tasks while improving speed.
Gradient Boosting Machine (GBM) is an extremely powerful supervised learning algorithm that is widely used in practice. GBM routinely features as a leading algorithm in machine learning competitions such as Kaggle and the KDDCup. In this work, we propose Accelerated Gradient Boosting Machine (AGBM) by incorporating Nes…
A method makes particle filters differentiable without altering their forward pass.
problem Compatibility issues between particle filters and automatic differentiation.
method Introduces a correction to particle weights using the stop-gradient operator.
result Automatic differentiation produces good estimators for gradients and second-order derivatives.
Characterizes problems solvable via linear convergence algorithms.
problem Optimization problems solvable with linear convergence.
method Riemannian gradient descent.
result Characterized problems solvable via linear convergence.
Quantum field theory connects deep neural networks to criticality.
problem Understanding the criticality and training dynamics of deep neural networks.
method Constructing quantum field theory for deep neural networks, computing corrections to correlation functions.
result Found precise analogy with O(N) vector model, providing corrections to correlation length. DAIS improves AIS for differentiable marginal likelihood estimation.
problem Differentiable marginal likelihood estimation for complex models.
method Proposes Differentiable Annealed Importance Sampling (DAIS) to make AIS differentiable.
result DAIS achieves convergence and consistency in Bayesian linear regression.
UCPO improves diversity in reinforcement learning models, maintaining high accuracy.
problem RLVR objectives often lead to diversity collapse, reducing coverage of correct solutions.
method UCPO adds a conditional uniformity penalty to GRPO, redistributing probability mass.
result UCPO improves Pass@K and diversity while maintaining competitive Pass@1 accuracy.
SGLRW improves robustness of stochastic gradient MCMC methods.
problem Sensitivity to minibatch size and gradient noise in stochastic-gradient MCMC methods.
method Proposes Stochastic Gradient Lattice Random Walk (SGLRW) with lattice-based discretization.
result SGLRW remains stable in regimes where SGLD fails, including heavy-tailed gradient noise.
Introduces TDRC to balance TD's ease and soundness.
problem TD learning's instability and divergence issues.
method Gradient Temporal-Difference Learning with Regularized Corrections (TDRC).
result TDRC performs as well as TD when TD works, but is sound in divergent cases.
This research improves online learning by correcting for target shift in machine learning.
problem Online learning struggles with distributional shift, especially in target values.
method Derives closed-form expressions for online and offline learning, and target correction.
result Online kernel-based learning can learn the same predictor as offline learning with target correction.
When samples have internal structure, we often see a mismatch between the objective optimized during training and the model's goal during inference. For example, in sequence-to-sequence modeling we are interested in high-quality translated sentences, but training typically uses maximum likelihood at the word level. The…
Variational inference transforms posterior inference into parametric optimization thereby enabling the use of latent variable models where otherwise impractical. However, variational inference can be finicky when different variational parameters control variables that are strongly correlated under the model. Traditiona…
Rescaled ASGD optimizes distributed learning under heterogeneous data.
problem Vanilla ASGD biases towards a frequency-weighted average of local objectives.
method Rescale worker stepsizes by their computation times.
result Rescaled ASGD converges to the correct global objective in fixed-computation model.
Stein variational gradient decent (SVGD) has been shown to be a powerful approximate inference algorithm for complex distributions. However, the standard SVGD requires calculating the gradient of the target density and cannot be applied when the gradient is unavailable. In this work, we develop a gradient-free variant …
Proposes EDM algorithm to accelerate model training in distributed networks.
problem Hindered effectiveness of distributed stochastic optimization algorithms due to data heterogeneity and network sparsity.
method Introduces Exact-Diffusion with Momentum (EDM) algorithm, incorporating momentum techniques to mitigate bias and enhance convergence rate.
result EDM algorithm converges sub-linearly to the optimal solution, radius independent of data heterogeneity, for non-convex objective functions.
Unified framework for decentralized bilevel optimization with various heterogeneity-correction strategies.
problem Decentralized bilevel optimization with neighborhood communications and data heterogeneity.
method SPARKLE: Single-loop Primal-dual Algorithm for decentralized bilevel optimization, incorporating various heterogeneity-correction techniques.
result Unified convergence analysis for SPARKLE with state-of-the-art convergence rates compared to existing algorithms.
Study on test risk dynamics in learning theory with stochastic gradient flow.
problem Understanding test risk in stochastic gradient flow dynamics.
method Path integral formulation for small learning rates, explicit computation for weak features.
result Explicit corrections due to stochastic term in dynamics, good agreement with simulations.
P3I learns holistic scene representations from a single image.
problem Inferring camera poses, object locations, and global scene structures from a single image.
method Combines search-based and gradient-based algorithms.
result P3I outperforms baselines on various image manipulation tasks.
We give a rigorous analysis of the statistical behavior of gradients in a randomly initialized fully connected network N with ReLU activations. Our results show that the empirical variance of the squares of the entries in the input-output Jacobian of N is exponential in a simple architecture-dependent constant beta, gi…
New method corrects bias in feature importance measures of GBM.
problem Bias in feature importance measures of GBM.
method Cross-validated unbiased base learners.
result Significant improvement in feature importance measures with minimal computational cost.
SupSup model learns thousands of tasks without forgetting, using randomly initialized subnetworks.
problem Sequentially learning many tasks without forgetting.
method Randomly initialized base network with task-specific subnetworks (supermasks).
result Gradient-based optimization can identify the correct subnetwork for new tasks.
Corrects errors in ILA for Bayesian inference in LGMs.
problem Error in ILA for non-Gaussian likelihoods in LGMs.
method Importance sampling scheme to correct ILA errors.
result Corrected posterior converges to the true posterior with increased samples.