The paper analyzes how forgetting in LLMs is linked to simple task-upstream example associations.
problem Forgetting of upstream knowledge in fine-tuned LLMs.
method Empirical analysis of forgotten examples in N upstream examples after M new tasks, using low-rank matrix approximation. result Forgetting can be predicted efficiently using matrix completion over empirical associations.
Model forgets examples; this research predicts which ones to replay.
problem Language models forget examples during updates, leading to errors.
method Train forecasting models to predict which examples will be forgotten.
result Forecasting models can reduce forgetting of upstream pretraining examples.
The paper finds a surprising positive correlation between upstreamness and downstreamness in global value chains.
problem The puzzling positive correlation between upstreamness and downstreamness in industries and countries.
method Analysis of a simple model of random Input/Output tables and experiments on empirical data.
result Upstreamness and downstreamness of the same industrial sector/country are positively correlated with a slope close to +1.
Study uses multidimensional SE-NBD process to analyze default portfolios and identify shock amplification.
problem Analyzing interactions and shock propagation in default portfolios with multiple sectors.
method Applied multidimensional self-exciting negative binomial distribution (SE-NBD) process to 13 sectors.
result Identified upstream and downstream sectors, showing shock amplification in default portfolios.
NeVI-Cut uses neural networks to efficiently propagate uncertainty without feedback.
problem Efficiently propagating uncertainty in downstream Bayesian analysis without feedback.
method NeVI-Cut combines neural networks and normalizing flows for variational inference.
result NeVI-Cut achieves significant computational gains and higher accuracy than traditional methods.
UBM transfers bias mitigation from upstream to downstream tasks efficiently.
problem Bias in fine-tuned language models across various tasks.
method Apply bias mitigation to an upstream model, then fine-tune a downstream model on this mitigated model.
result UBM effects transfer to new downstream tasks, creating less biased models.
Paper examines adversarial attacks on weather forecasting models, focusing on TC trajectory prediction.
problem Adversarial attacks can mislead downstream TC trajectory predictions in DLWF models.
method Proposes Cyc-Attack, a method using a surrogate model and skewness-aware loss function to generate adversarial TC paths.
result Cyc-Attack achieves higher true positive rates and lower false alarm rates compared to conventional methods.
Study shows scaling up models doesn't always improve downstream tasks.
problem Understanding why scaling up models doesn't always improve downstream performance.
method Systematic study of 4800 experiments on various models, analyzing performance on 20 downstream tasks.
result Performance on downstream tasks saturates as model size increases, revealing a nonlinear relationship.
New algorithm REFUEL shows multitask representation learning is more sample-efficient in RL.
problem Understanding the benefit of representation learning in reinforcement learning.
method Developed REFUEL algorithm for multitask low-rank RL, analyzing both upstream and downstream tasks.
result Multitask representation learning is provably more sample-efficient than individual task learning.
Supply chains lend themselves to blockchain technology, but certain challenges remain, especially around invoice financing. For example, the further a supplier is removed from the final consumer product, the more difficult it is to get their invoices financed. Moreover, for competitive reasons, retailers and manufactur…
Deep tabular models outperform GBDT in medical diagnosis tasks.
problem Transfer learning for tabular data in medical diagnosis.
method Proposes a pseudo-feature method for transfer learning between different feature sets.
result Tabular neural networks outperform GBDT in medical diagnosis tasks.
New methods detect modular structure in neural networks, revealing surprising effects of dropout.
problem Detecting functional modules in neural networks for learning, compositionality, and generalization.
method Two families of methods: upstream and downstream, to define similarity between units.
result Dropout dramatically increased modularity, and there's little agreement between upstream and downstream methods.
This paper investigates multi-task reinforcement learning in non-Markovian decision making, showing benefits in sample efficiency.
problem Investigating multi-task reinforcement learning in non-Markovian decision making processes.
method Developed a joint model class for tasks and used the η-bracketing number to quantify complexity and similarity. result Multi-task reinforcement learning can improve sample efficiency in non-Markovian decision making processes.
ASCEND discovers causal relationships in multi-omics data by leveraging known hierarchical structure.
problem Causal inference in high-dimensional multi-omics data, especially when ignoring the hierarchical structure.
method Two-tiered divide-and-conquer strategy with ancestral conditioning sets.
result Achieves polynomial-time complexity and accurately recovers ancestral relationships.
Enhances performance on downstream tasks using multi-domain data.
problem Improving performance on tasks with limited downstream data.
method Deep transfer learning framework leveraging shared and domain-specific features.
result Significantly improves convergence rate for learning Lipschitz functions.
We study the problem of learning representations with controllable connectivity properties. This is beneficial in situations when the imposed structure can be leveraged upstream. In particular, we control the connectivity of an autoencoder's latent space via a novel type of loss, operating on information from persisten…
CARV reduces compute cost for downstream pipelines using diffusion models.
problem High variance in Monte Carlo estimators from diffusion models limits compute efficiency.
method CARV uses hierarchical MC estimation with amortized upstream computation and stratified-inverse-CDF.
result CARV delivers 2-3x effective compute multipliers without changing the objective.
Simplex-valued data appear throughout statistics and machine learning, for example in the context of transfer learning and compression of deep networks. Existing models for this class of data rely on the Dirichlet distribution or other related loss functions; here we show these standard choices suffer systematically fr…
Study strategic competition in commodity markets using impulse-switching controls.
problem Strategic competition between upstream and downstream firms in commodity markets.
method Non-zero-sum stochastic differential game with mixed impulse/switching controls.
result Multiple Nash equilibria found, depending on the number of switches by the downstream firm.
New method ranks sectors and countries using local and aggregate I-O data.
problem Ranking sectors and countries in global value chains using incomplete I-O tables.
method Rank-1 approximation to I-O tables using local and aggregate information. result Consistently good performance in reconstructing rankings of upstreamness and downstreamness.
This paper emphasizes model transparency and interpretation in insurance.
problem Ensuring models do not discriminate and are explainable.
method Exploring tools to control actuarial models using machine learning.
result Interpretability methods can adapt explanations to different audiences.
Currently, progressively larger deep neural networks are trained on ever growing data corpora. As this trend is only going to increase in the future, distributed training schemes are becoming increasingly relevant. A major issue in distributed training is the limited communication bandwidth between contributing nodes o…
Network models assume unrealistic idiosyncratic risk, which can be mitigated by allowing for correlated shocks.
problem Network models assume idiosyncratic risk, which can be unrealistic and lead to incorrect predictions.
method Proposed a production-based asset pricing model to account for substitutability between trade partners and correlation in supply and demand shocks.
result Assets positively exposed to average propagation of upstream and downstream shocks earn lower average risk premia.
Optimizes ad pruning in sponsored search systems using reinforcement learning.
problem How to efficiently select top K ads from N candidates to maximize revenue.
method Model-free reinforcement learning approach considering downstream as a black-box environment.
result Remarkable improvements in revenue achieved through reinforcement learning.
Reinforcement learning algorithms are known to be sample inefficient, and often performance on one task can be substantially improved by leveraging information (e.g., via pre-training) on other related tasks. In this work, we propose a technique to achieve such knowledge transfer in cases where agent trajectories conta…
A new method corrects bias in machine learning for trading by filtering out non-executable prices.
problem Contaminated data from price limits causes misleading performance metrics in trading models.
method Mask-first design with tradability filters, GPU optimization, and penalized losses.
result The system achieves Sharpe ratios of 1.63 on real data, significantly outperforming existing methods.
Study examines value relevance of oil and gas reserve disclosures in London Stock Exchange.
problem Uncertainty in oil and gas reserves poses accounting challenges for investors.
method Empirical analysis using archival data and multifactor framework.
result Changes in reserves and their components are associated with share returns, but insignificantly due to oil price and longitudinal effects. Quality of disclosures positively impacts share returns.
In recent years, the advances in single-cell RNA-seq techniques have enabled us to perform large-scale transcriptomic profiling at single-cell resolution in a high-throughput manner. Unsupervised learning such as data clustering has become the central component to identify and characterize novel cell types and gene exp…
The current leading computer vision models are typically feed forward neural models, in which the output of one computational block is passed to the next one sequentially. This is in sharp contrast to the organization of the primate visual cortex, in which feedback and lateral connections are abundant. In this work, we…
We propose a practical approach based on federated learning to solve out-of-domain issues with continuously running embedded speech-based models such as wake word detectors. We conduct an extensive empirical study of the federated averaging algorithm for the "Hey Snips" wake word based on a crowdsourced dataset that mi…
We model a network economy with three sectors: downstream firms, upstream firms, and banks. Agents are linked by productive and credit relationships so that the behavior of one agent influences the behavior of the others through network connections. Credit interlinkages among agents are a source of bankruptcy diffusion…
New approach speeds up optimization with repeated gradient steps on same batch.
problem Performance bottlenecks in massive parallel pipelines with large batch sizes.
method Data echoing, taking repeated gradient steps on the same batch.
result Data echoing affords speedups on curvature-dominated part of convergence rate.
LIM enhances investment performance and efficiency at scale.
problem Diminishing returns and rising labor/time costs in traditional quantitative investment research.
method End-to-end learning and universal modeling to create a global patterns foundation model.
result Optimized performance for specific tasks through transfer learning of global patterns.
ACL improves robustness with unlabeled data, and we analyze its generalization using Rademacher complexity.
problem Improving robustness of deep networks against adversarial attacks using unlabeled data.
method We analyze the generalization performance of Adversarial Contrastive Learning (ACL) using Rademacher complexity.
result The average adversarial risk of the downstream tasks can be upper bounded by the adversarial unsupervised risk of the upstream task.
Ensembling word embeddings to improve distributed word representations has shown good success for natural language processing tasks in recent years. These approaches either carry out straightforward mathematical operations over a set of vectors or use unsupervised learning to find a lower-dimensional representation. Th…
Improved neural topic model for semi-supervised learning.
problem Representing textual data in an interpretable manner with limited labeled data.
method Label-Indexed Neural Topic Model (LI-NTM) that combines deep generative models with semi-supervised learning.
result LI-NTM outperforms existing models in document reconstruction and classifier performance.
With recent progress in algorithms and the availability of massive amounts of computation power, application of machine learning techniques is becoming a hot topic in the oil and gas industry. One of the most promising aspects to apply machine learning to the upstream field is the rock facies classification in reservoi…
HydroNets use river structure to improve hydrologic predictions.
problem Scalable and accurate hydrologic models are needed for climate change impacts.
method HydroNets are deep neural networks that incorporate river network structure.
result HydroNets improve predictions with fewer data, especially at longer horizons.
Recurrent convolution (RC) shares the same convolutional kernels and unrolls them multiple steps, which is originally proposed to model time-space signals. We argue that RC can be viewed as a model compression strategy for deep convolutional neural networks. RC reduces the redundancy across layers. However, the perform…
Proposes a method to construct risk-neutral marginals from arbitrage-free option prices.
problem Lack of risk-neutral marginals that are free of arbitrage and easy to use.
method Explicit construction of risk-neutral marginals from discrete arbitrage-free option prices.
result Explicit construction guarantees risk-neutral marginals free of butterfly and calendar arbitrage.
This research smooths out fluid equations to avoid sudden shocks.
problem Formation of shock singularities in compressible fluid equations.
method Information geometric regularization of unidimensional pressureless Euler equations.
result Smooth global solutions without artificial viscosity.
Multi-agent coordination is prevalent in many real-world applications. However, such coordination is challenging due to its combinatorial nature. An important observation in this regard is that agents in the real world often only directly affect a limited set of neighbouring agents. Leveraging such loose couplings amon…
Neuronal circuits formed in the brain are complex with intricate connection patterns. Such complexity is also observed in the retina as a relatively simple neuronal circuit. A retinal ganglion cell receives excitatory inputs from neurons in previous layers as driving forces to fire spikes. Analytical methods are requir…
The paper tackles fair representation learning by smoothing feature mappings.
problem Legal liability for discriminatory use of data by organizations.
method Mapping features to a fair representation space, certifying fairness through chi-squared mutual information.
result Smoothing representation distribution provides generalization guarantees of fairness and maintains accuracy for downstream tasks.
Method adapts frozen models for few-shot tasks without training.
problem Deployment constraints limit model updates, necessitating new adaptation methods.
method Exponential tilting of latent distribution for inference.
result Method outperforms parameter-update methods across benchmarks.
A new conformal prediction framework for two-stage models identifies stage-wise uncertainty.
problem Limited coverage guarantees and lack of modular structure understanding in existing conformal prediction methods.
method Decomposes prediction residuals into stage-specific components, calibrates parameters using FWER control, and adapts to non-stationary settings.
result Improves coverage and identifies stage-wise error contributions compared to standard conformal methods.
Paper improves molecular property prediction using denoising autoencoders.
problem Limited data for molecular property prediction from 3D structures.
method Pre-training via denoising for learning molecular force fields.
result Achieves new state-of-the-art performance on QM9 dataset.
Advanced travel information and warning, if provided accurately, can help road users avoid traffic congestion through dynamic route planning and behavior change. It also enables traffic control centres mitigate the impact of congestion by activating Intelligent Transport System (ITS) proactively. Deep learning has beco…