The paper proposes a learning algorithm that improves adaptability and generalization.
problem Improving adaptability and generalization in learning models.
method Learning to meta-learn by meta-finetuning on related tasks before adapting to specific tasks.
result Learning to meta-learn improves adaptability and generalization across various tasks.
Framework adapts to new tasks based on prior knowledge.
problem Models struggle to adapt to novel tasks without direct experience.
method Learned task representations and meta-mappings to transform them.
result Meta-mapping achieves 80-90% performance on novel tasks.
New RL algorithms adapt to time limits, improving task performance.
problem Fixed RL behaviors cannot adapt to different time restrictions.
method Introduced two algorithms for time adaptive RL: Independent Gamma-Ensemble and n-Step Ensemble.
result Zero-shot adaptation between different time restrictions.
AvaGrad optimizes vision tasks by decoupling learning rate and adaptability.
problem Improving optimization methods for vision tasks.
method Derives AvaGrad, a new optimizer that decouples learning rate and adaptability.
result AvaGrad outperforms SGD on vision tasks when adaptability is properly tuned.
Adaptive methods learn from multiple datasets, leveraging similarities and robust to outliers.
problem Simultaneously analyze multiple datasets with possible similarities and differences.
method Adaptive multi-task learning methods that automatically utilize similarities and handle differences.
result Sharp statistical guarantees and robustness against outlier tasks demonstrated.
M-L2O adapts fast to new tasks by self-adapting during test-time.
problem L2O optimizers struggle with out-of-distribution tasks.
method Meta-training an L2O optimizer to adapt quickly to new tasks.
result M-L2O converges significantly faster than vanilla L2O with only 5 steps of adaptation.
FAST improves fast and stable task adaptation in DNNs.
problem Catastrophic forgetting in fine-tuned pretrained models.
method Introducing FAST, an easy-to-implement fine-tuning algorithm.
result FAST learns target tasks faster and retains source knowledge longer.
CoLoRA leverages task similarity to boost fine-tuning efficiency.
problem Efficiently fine-tuning large foundation models with scarce labeled data.
method CoLoRA trains a shared adapter for task similarity and personalized adapters for user-specific tasks.
result CoLoRA significantly boosts fine-tuning performance when tasks are similar.
FLAP adapts policies quickly to new tasks using shared linear representations.
problem Adapting policies to new tasks efficiently and effectively.
method FLAP uses a shared linear representation and a separate adapter network for quick adaptation.
result FLAP achieves up to 8X faster adaptation and significantly better performance on out-of-distribution tasks.
Adversarial weighting improves regression task adaptation.
problem Improving regression performance across domains with covariate shift.
method Adversarial network algorithm for instance weighting and task learning.
result The method enhances regression accuracy on target domains.
FLoE adapts LLMs by selectively deploying LoRA adapters based on layer importance and task requirements.
problem Uniform LoRA deployment across all layers leads to inefficient and redundant parameter allocation.
method FLoE uses Fisher information to dynamically identify task-critical layers and optimizes LoRA ranks.
result FLoE achieves significant efficiency-accuracy trade-offs, especially in resource-constrained environments.
DPTA improves CIL by adapting PTMs with dual prototypes.
problem Catastrophic forgetting in incremental learning with pre-trained models.
method Dual-Prototype Network with Task-wise Adaptation (DPTA).
result DPTA consistently outperforms recent methods by 1\%-5\% on multiple benchmarks.
To train neural machine translation models simultaneously on multiple tasks (languages), it is common to sample each task uniformly or in proportion to dataset sizes. As these methods offer little control over performance trade-offs, we explore different task scheduling approaches. We first consider existing non-adapti…
Fine-tuning large pre-trained models is an effective transfer mechanism in NLP. However, in the presence of many downstream tasks, fine-tuning is parameter inefficient: an entire new model is required for every task. As an alternative, we propose transfer with adapter modules. Adapter modules yield a compact and extens…
BoRA finetunes multi-task LLMs by sharing information through hierarchical priors.
problem Limited data for some tasks in multi-task LLMs.
method Bayesian hierarchical low-rank adaption.
result BoRA outperforms individual and unified model approaches.
DANCE improves prediction set efficiency for deep learning models.
problem Inefficient, overly conservative prediction sets for pre-trained models.
method DANCE combines adaptive kernel regression and nearest-neighbor approach.
result DANCE produces more efficient and robust prediction sets.
Meta-learning improved by using information theory to prioritize data-driven adaptation.
problem Challenges in meta-learning due to the need for mutually-exclusive tasks.
method Designing a meta-regularization objective using information theory.
result Successfully uses data from non-mutually-exclusive tasks to efficiently adapt to novel tasks.
Adaptive data fusion boosts efficiency in multi-task optimization.
problem Multi-task non-smooth optimization in various fields.
method Adaptive data fusion approach leveraging commonalities among objectives.
result Significant improvements in sample efficiency with sharp statistical guarantees.
Model-Agnostic Meta-Learning (MAML) and its variants have achieved success in meta-learning tasks on many datasets and settings. On the other hand, we have just started to understand and analyze how they are able to adapt fast to new tasks. For example, one popular hypothesis is that the algorithms learn good represent…
APT-Gen generates tasks to help RL learn in hard problems.
problem Learning in hard exploration problems.
method APT-Gen uses a task generator to create tasks from a parameterized space, balancing performance and similarity to target tasks.
result APT-Gen outperforms baselines in grid world and robotic manipulation tasks.
Gradient-based meta-RL fails with incorrect task distributions, leading to instability and poor performance.
problem Gradient-based meta-RL's sensitivity to task distributions causes instability and poor performance.
method Proposes meta Active Domain Randomization (meta-ADR) to learn task distributions for gradient-based meta-RL.
result Meta-ADR improves stability and generalization of MAML on simulated locomotion and navigation tasks.
Paper introduces RAS for robust MTL with contamination.
problem Representation-based multi-task learning struggles with contamination.
method Robust and Adaptive Spectral (RAS) method.
result RAS prevents negative transfer and performs well with up to 80% contamination.
Bayesian approach adapts deep network structure for continual learning.
problem Training neural networks with sequential or streaming tasks.
method Bayesian approach to learn deep network structure for each task.
result Model performs comparably or better than recent advances in continual learning.
We introduce the Adaptive Skills, Adaptive Partitions (ASAP) framework that (1) learns skills (i.e., temporally extended actions or options) as well as (2) where to apply them. We believe that both (1) and (2) are necessary for a truly general skill learning framework, which is a key building block needed to scale up t…
BI-MAML learns multiple tasks without forgetting old ones.
problem Catastrophic forgetting in meta learning.
method Incremental model adaptation with balanced learning strategy.
result BI-MAML outperforms state-of-the-art models in accuracy and efficiency.
How can deep learning systems flexibly reuse their knowledge? Toward this goal, we propose a new class of challenges, and a class of architectures that can solve them. The challenges are meta-mappings, which involve systematically transforming task behaviors to adapt to new tasks zero-shot. The key to achieving these c…
Study compares adaptive vs fixed query learning methods.
problem Comparing adaptive and fixed query learning methods for task approximation.
method Examined in-context and agentic learning in two settings: unrestricted and realizable.
result Adaptivity does not hinder performance in unrestricted setting but can in realizable setting.
CLAW adapts weights to balance continual learning and catastrophic forgetting.
problem Balancing continual learning across multiple tasks without forgetting previous knowledge.
method Probabilistic modelling and variational inference to adaptively share network components.
result CLAW achieves state-of-the-art performance in continual learning benchmarks.
PDTS improves robustness in sequential decision-making.
problem Robust active task sampling for efficient and reliable decision-making.
method Characterizes robust active task sampling as a Markov decision process, proposes PDTS method.
result Significantly improves zero-shot and few-shot adaptation robustness.
Paper analyzes convergence of continual learning with adaptive methods.
problem Preventing catastrophic forgetting in sequential learning tasks.
method Adaptive method for nonconvex continual learning (NCCL) adjusts step sizes of previous and current tasks.
result Proposed adaptive method achieves same convergence rate as SGD when catastrophic forgetting is suppressed.
Meta-SAGE improves deep RL scalability for CO tasks by adapting pre-trained models to larger-scale problems.
problem Improving scalability of deep reinforcement learning models for combinatorial optimization tasks.
method Meta-SAGE combines a scale meta-learner and scheduled adaptation with guided exploration to adjust model parameters for larger-scale problems.
result Meta-SAGE outperforms previous methods and significantly improves scalability in CO tasks.
Generative Adapter adapts LMs with a single forward pass, reducing inference overhead.
problem Efficient adaptation of large language models for new contexts.
method Generative Adapter directly maps new contexts to low-rank LM adapters via self-supervised learning.
result Significant reduction in inference overhead with no need for fine-tuning.
Deep reinforcement learning agents have recently been successful across a variety of discrete and continuous control tasks; however, they can be slow to train and require a large number of interactions with the environment to learn a suitable policy. This is borne out by the fact that a reinforcement learning agent has…
The paper tackles learning from similar but not identical linear representations, improving performance over single-task learning.
problem Understanding how to learn from tasks with similar but not exactly the same linear representations, especially when dealing with outlier tasks.
method Proposes adaptive and robust penalized empirical risk minimization and spectral methods.
result Both methods outperform single-task learning when representations are similar and perform at least as well otherwise, with minimax optimality demonstrated.
Gradient-based meta-learners such as MAML are able to learn a meta-prior from similar tasks to adapt to novel tasks from the same distribution with few gradient updates. One important limitation of such frameworks is that they seek a common initialization shared across the entire task distribution, substantially limiti…
SAP learns efficient task-specific parameter subspaces for few-shot learning.
problem Efficient few-shot learning with limited data.
method Subspace Adaptation Prior (SAP) learns task-specific parameter subspaces for efficient few-shot learning.
result SAP yields superior or competitive performance in few-shot image classification.
Improves unsupervised domain adaptation by enforcing feature extractor to focus on task-relevant information.
problem Leveraging label information from source domain for accurate target domain models without labels.
method Variational Information Bottleneck (VBDA) method that explicitly enforces feature extractor to ignore irrelevant task factors.
result Significantly outperforms state-of-the-art methods across three domain adaptation benchmark datasets.
Transformer learns to infer partial MDPs for efficient in-context adaptation and exploration.
problem Efficiently adapt and explore in-context without gradient-based updates.
method Uses a transformer to learn inference from training tasks, considering hypothesis space of partial models.
result Adaptation speed and exploration-exploitation balance approach those of an exact posterior sampling oracle.
Empirical study shows consistent meta-RL algorithms adapt to OOD tasks.
problem Theoretical consistency of meta-RL algorithms and its practical implications.
method Empirical investigation of representative meta-RL algorithms, focusing on consistency and adaptation to out-of-distribution tasks.
result Theoretical consistent algorithms can adapt to OOD tasks, while inconsistent ones cannot, but can still fail for poor exploration.
The capacity of meta-learning algorithms to quickly adapt to a variety of tasks, including ones they did not experience during meta-training, has been a key factor in the recent success of these methods on few-shot learning problems. This particular advantage of using meta-learning over standard supervised or reinforce…
A new meta-learning BO approach that bypasses surrogate models and directly learns task utility.
problem Scalability issues and sensitivity to task similarity in existing meta-learning BO methods.
method Directly learns the utility of queries across tasks, models task uncertainty, and includes an auxiliary model for robust adaptation.
result Demonstrates strong anytime performance and outperforms state-of-the-art methods in various benchmarks.
MetaVRF learns adaptive kernels for fast few-shot learning.
problem Few-shot learning with limited data.
method MetaVRF with latent variable model and variational inference.
result MetaVRF produces kernels with high representational power and fast adaptation.
The aim of multi-task reinforcement learning is two-fold: (1) efficiently learn by training against multiple tasks and (2) quickly adapt, using limited samples, to a variety of new tasks. In this work, the tasks correspond to reward functions for environments with the same (or similar) dynamical models. We propose to l…
Crowdsourcing platforms provide marketplaces where task requesters can pay to get labels on their data. Such markets have emerged recently as popular venues for collecting annotations that are crucial in training machine learning models in various applications. However, as jobs are tedious and payments are low, errors …
Proposes Ada-Sit method for mortality prediction of rare diseases.
problem Data insufficiency and clinical diversity of rare diseases make mortality prediction hard.
method Initialization-sharing multi-task learning method (Ada-Sit) for fast adaptation to similar tasks.
result Experimental results show the proposed model is effective for mortality prediction of diverse rare diseases.
Online learning with streaming data in a distributed and collaborative manner can be useful in a wide range of applications. This topic has been receiving considerable attention in recent years with emphasis on both single-task and multitask scenarios. In single-task adaptation, agents cooperate to track an objective o…
TASFAR adapts regression models without labeled source data.
problem Lack of labeled source data for domain adaptation.
method Uses prediction confidence to estimate target label distribution and calibrate source model.
result Substantially reduces errors in various regression tasks.
Improves model calibration and selection in unsupervised domain adaptation.
problem Distribution shifts in unsupervised domain adaptation.
method Developed a novel importance weighted group accuracy estimator.
result Improves state-of-the-art performances by 22% in model calibration and 14% in model selection.