nGPT learns to transfer learning rates across model dimensions and token horizons.
problem nGPT does not transfer learning rates across model size and token horizon.
method Combining numerical experiments with alignment exponents, a novel nGPT parameterization νGPT is developed.
result νGPT exhibits learning rate transfer across width, depth, and token horizon.
Proof of learning rate transfer in MLPs with μ μ μ P parameterization.
problem Understanding and optimizing learning rates in neural networks with different parameterizations.
method Theoretical analysis and empirical validation of learning rate transfer in MLPs with μ μ μ P, SP, and NTP parameterizations. result The optimal learning rate converges to a non-zero constant as width goes to infinity under μ μ μ P, explaining learning rate transfer. Flexible framework for transfer learning with optimal rates.
problem Inference about a target population using related source data.
method Adaptive transfer learning framework allowing covariate-dependent relationships.
result Achieves minimax optimal rates of convergence by adapting to transfer relationship.
This paper quantifies hyperparameter transfer and finds embedding layer learning rate is key.
problem Quantifying optimal hyperparameters for large language models across scales.
method Developed three metrics to quantify hyperparameter transfer and investigated the importance of embedding layer learning rate.
result Maximal Update (μP) parameterization offers high-quality learning rate transfer compared to standard parameterization (SP).
Transfer learning improves chaotic dynamics predictions with less data.
problem Efficiently predicting chaotic dynamics with limited data.
method Transfer learning for nonlinear dynamics, optimizing transfer rate and leveraging small-scale turbulence universality.
result Significantly more accurate inference of chaotic dynamics achieved.
We present a new approach for transferring knowledge from groups to individuals that comprise them. We evaluate our method in text, by inferring the ratings of individual sentences using full-review ratings. This approach, which combines ideas from transfer learning, deep learning and multi-instance learning, reduces t…
Study examines the impact of target data in transfer learning.
problem Understanding the value of target data in transfer learning.
method Established minimax-rates for source and target sample sizes, introduced transfer exponents.
result Performance limits in transfer learning are captured by transfer exponents.
Q-learning is one of the most popular methods in Reinforcement Learning (RL). Transfer Learning aims to utilize the learned knowledge from source tasks to help new tasks to improve the sample complexity of the new tasks. Considering that data collection in RL is both more time and cost consuming and Q-learning converge…
Learning from small amounts of labeled data is a challenge in the area of deep learning. This is currently addressed by Transfer Learning where one learns the small data set as a transfer task from a larger source dataset. Transfer Learning can deliver higher accuracy if the hyperparameters and source dataset are chose…
New study reveals surprising adaptive rates in model selection for transfer learning.
problem Model selection in transfer learning with varying transfer distances.
method Theoretical analysis focusing on classification.
result Adaptive rates can be arbitrarily slower than oracle rates.
AdaTrans adapts to feature and sample transfer in high-dimensional regression.
problem High-dimensional linear regression with more features than samples.
method F-AdaTrans and S-AdaTrans methods using fused-penalties and adaptive weights.
result AdaTrans achieves convergence rates close to oracle estimators and near-minimax optimal rates.
Develops methods to measure and set function-space learning rates in neural networks.
problem Measuring and optimizing changes in neural network output functions.
method Efficient methods to measure and set function-space learning rates, requiring minimal computational overhead.
result Demonstrates FLeRM (Function-space Learning Rate Matching) for hyperparameter transfer across model scales.
New method tackles concept shifts in nonparametric regression using robust and adaptive transfer learning.
problem Concept shifts and sample scarcity in target domains hinder nonparametric regression.
method Robust and adaptive transfer learning procedure leveraging fixed bandwidth Gaussian kernels.
result Spectral algorithms with fixed bandwidth Gaussian kernels attain minimax convergence rates for nonparametric regression.
This paper improves kernel-based regression using transfer learning.
problem Improving generalization performance in kernel-based regression.
method Two-step kernel-based estimator for known transferable sources and novel aggregation algorithm for unknown sources.
result Established statistical properties and validated effectiveness of proposed methods.
Cross-domain recommendation has long been one of the major topics in recommender systems. Recently, various deep models have been proposed to transfer the learned knowledge across domains, but most of them focus on extracting abstract transferable features from auxilliary contents, e.g., images and review texts, and th…
We propose to apply deep transfer learning from computer vision to static malware classification. In the transfer learning scheme, we borrow knowledge from natural images or objects and apply to the target domain of static malware detection. As a result, training time of deep neural networks is accelerated while high c…
Paper proposes a transfer learning method for improving matrix completion.
problem Improving estimation of a low-rank target matrix using auxiliary data.
method Transfer learning procedure leveraging prior information on favorable source datasets.
result Method outperforms traditional methods when source datasets are close to the target matrix.
Study shows benefits of transfer learning with neural networks.
problem Understanding generalization errors in transfer learning.
method Mean-field analysis applied to α α α -ERM and fine-tuning. result Established conditions for generalization error and convergence rates.
Estimates target GGM using auxiliary studies with false discovery rate control.
problem Estimating high-dimensional GGMs from related studies.
method Transfer learning with Trans-CLIME and debiased Trans-CLIME estimators.
result Debiased Trans-CLIME estimator provides element-wise asymptotic normality and false discovery rate control.
The paper uses thermodynamics to improve machine learning representation quality.
problem Improving the quality of learned representations for transfer learning.
method Formal connection with thermodynamics, iso-classification process, traversing the equilibrium surface.
result Demonstrates how to transfer representations while keeping classification loss constant.
Human learners have the natural ability to use knowledge gained in one setting for learning in a different but related setting. This ability to transfer knowledge from one task to another is essential for effective learning. In this paper, we study transfer learning in the context of nonparametric classification based …
Paper tackles transfer learning for contextual multi-armed bandits under covariate shift.
problem Nonparametric contextual multi-armed bandits with covariate shift.
method Established minimax rate of convergence, proposed transfer learning algorithm.
result Achieved near-optimal statistical guarantees for learning in target domain.
Localized transfer learning improves nonparametric regression performance.
problem Improving nonparametric regression performance on target tasks.
method Localized transfer learning framework that models heterogeneity and partition covariate space into cells.
result Sharp minimax rates show local transfer mitigates the curse of dimensionality.
A simple strategy prevents negative transfer in transfer learning.
problem Negative transfer in transfer learning where source representations harm target performance.
method Residual feature integration with a trainable target-side encoder.
result The method provably prevents negative transfer with theoretical guarantees.
Transfer learning for bandits with latent Lipschitz continuity.
problem Learning to transfer structural information from prior tasks to new tasks.
method Proposes a framework to estimate Lipschitz constant from prior tasks and apply it to new tasks.
result Regret bound close to oracle algorithm with full knowledge of Lipschitz constant under mild assumptions.
The paper improves high-dimensional linear regression prediction and estimation using auxiliary samples.
problem Estimating and predicting high-dimensional linear regression models with auxiliary samples.
method Proposes Trans-Lasso for data-driven transfer learning, establishing optimality for prediction and estimation.
result Knowledge from auxiliary samples can improve learning performance in target problems.
New method for RL tasks transfer using Lipschitz continuity.
problem Knowledge transfer in RL tasks over time.
method Established Lipschitz continuity between MDPs and applied it to RL.
result Improved convergence rate and no negative transfer with high probability.
Paper tackles federated learning with privacy, enhancing target data analysis.
problem Heterogeneity and privacy of distributed data in federated learning.
method Formulates federated differential privacy, studies statistical problems under privacy constraints.
result Federated differential privacy offers a balance between privacy and knowledge transfer.
Improves functional linear regression with shape transfer learning.
problem Data scarcity in functional linear models.
method Shape-based transfer learning from auxiliary to target domains.
result Enhances robustness and generalizability of functional linear models.
Collaborative filtering (CF) aims to predict users' ratings on items according to historical user-item preference data. In many real-world applications, preference data are usually sparse, which would make models overfit and fail to give accurate predictions. Recently, several research works show that by transferring k…
We introduce a new weight-decay scaling rule to maintain sublayer gains across different widths in modern scale-invariant architectures.
problem In modern scale-invariant architectures, training quickly enters a steady state where normalization layers create backward scale sensitivity, degrading learning-rate transfer.
method We introduce a weight-decay scaling rule for AdamW that preserves sublayer gain across widths by equalizing the effective learning rate.
result Our empirical weight-decay scaling rule λ 2 ∝ d λ_2\propto \sqrt{d} λ 2 ∝ d approximately keeps sublayer gains width invariant, enabling zero-shot transfer of learning rate and weight decay. New optimizers control network width scaling, improving stability and transfer across different model sizes.
problem Designing stable optimizers for networks of varying widths.
method Interpreting optimizers as steepest descent under mean-normalized operator norms, enabling layerwise composability and width-independent bounds.
result New optimizers like row normalization and column normalization provide stable learning-rate transfer across different model widths.
TLRF improves timely COVID-19 outbreak detection with small sample size counties.
problem Balancing accuracy and speed in estimating COVID-19 case growth rates.
method Transfer Learning Random Forest (TLRF) framework for growth rate estimation.
result TLRF outperforms existing methods in predicting case growth rates and timely outbreak detection.
Study shows cliff-learning in transfer learning from foundation models.
problem Data-scaling of transfer learning from foundation models in low data regimes.
method Investigation of cliff-learning phenomenon through foundation-model analysis and toy models.
result Cliff-learning reflects compatibility between priors and tasks.
New method improves adversarial transferability from Bayesian neural networks.
problem Improving the effectiveness of black-box evasion attacks.
method Sampling from the posterior distribution of neural network weights to build a surrogate.
result Significantly improved success rates of state-of-the-art attacks (up to 83.2 percentage points).
Adaptive transfer learning model for varying mechanisms across domains.
problem Improving inference in a target domain by leveraging related source domains with varying mechanisms.
method Semi-parametric domain-varying coefficient model (DVCM) for structured transfer learning.
result Minimax rate-optimal adaptive transfer learning estimator with provable negative transfer safeguards.
MLR-SNet learns flexible LR schedules for diverse tasks.
problem Adapting LR schedules for non-convex optimization problems.
method Parameterizes LR schedules with an explicit mapping formulation.
result Meta-learned LR schedules are transferable and adaptable.
Paper tackles nonparametric classification with privacy constraints, achieving optimal accuracy.
problem Nonparametric classification under distributed differential privacy constraints.
method Minimax and adaptive transfer learning, considering privacy, sample sizes, and heterogeneity.
result Developed an adaptive classifier achieving optimal misclassification rate with privacy constraints.
Improves model accuracy in medical imaging with small datasets using transfer learning.
problem Challenges in training neural networks on small medical imaging datasets.
method Comparison of current techniques, proposing one cycle training, discriminative learning rates, gradual freezing, and parameter modification.
result Transfer learning is crucial for small datasets, especially when images from the same part of the body are available.
Bullseye Polytope improves clean-label poisoning attacks in transfer learning.
problem Poisoning neural networks with correctly labeled data.
method Creates poison images with centers close to target images in feature space.
result Improves attack success rate by 26.75% in end-to-end transfer learning.
SATL adapts to varying smoothness in hypothesis transfer learning.
problem Fixed kernel regularization fails in varying smoothness settings.
method Proposes SATL, a two-phase KRR algorithm with adaptive Gaussian kernels.
result SATL achieves minimax optimality with matching upper and lower bounds.
RIFLE improves deep transfer learning by reinitializing fully-connected layers.
problem Limited improvement in transfer learning accuracy with pre-trained models on small datasets.
method Re-Initializing fully-connected layers with random scratch during fine-tuning.
result Significant improvement in deep transfer learning accuracy across various datasets.
New algorithm improves knowledge transfer in dynamic decision-making.
problem Utilizing data from existing ventures to improve decision-making in new ventures.
method Proposes Transferred Fitted Q Q Q -Iteration algorithm for estimating optimal action-state function Q ∗ Q^* Q ∗ . result Significantly improved final learning error of Q ∗ Q^* Q ∗ function. The paper tackles transfer learning for growing matrix representations, improving estimation accuracy.
problem Structured matrix estimation under growing ambient dimensions and latent representations.
method Proposes a general transfer framework decomposing target parameters into embedded source components, low-rank innovations, and sparse edits. Develops an anchored alternating projection estimator.
result Establishes deterministic error bounds that separate target noise, representation growth, and source estimation error, yielding improved rates.
Framework improves target domain prediction using quantile matching.
problem Improving prediction accuracy in data-scarce target domains.
method Conditional quantile matching for distributional alignment.
result Empirical risk minimizer achieves tighter excess risk bound.
LGV boosts adversarial attacks by improving surrogate models.
problem Improving the transferability of black-box adversarial attacks.
method LGV uses a pretrained surrogate model and multiple weight sets from additional training epochs to generate an effective surrogate ensemble.
result LGV outperforms other test-time transformations by significant margins.
New method uses correlation-ratio for transfer learning, improving target model inference.
problem Difficulty in meeting similarity condition for transfer learning.
method Introduces linear correlation-ratio for building accurate relationships between models.
result Improves inference on target model by transferring information from source models.
Background: Pharmacokinetic evaluation is one of the key processes in drug discovery and development. However, current absorption, distribution, metabolism, excretion prediction models still have limited accuracy. Aim: This study aims to construct an integrated transfer learning and multitask learning approach for deve…