A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
Paper tackles catastrophic overfitting in single-step adversarial training.
Adversarial examples have become one of the largest challenges that machine learning models, especially neural network classifiers, face. These adversarial examples break the assumption of attack-free scenario and fool state-of-the-art (SOTA) classifiers with insignificant perturbations to human. So far, researchers ac…
Efficient regularization mitigates catastrophic overfitting in single-step adversarial training.
A new method reduces adversarial training time without overfitting.
We present an extension of our Molecular Transformer architecture combined with a hyper-graph exploration strategy for automatic retrosynthesis route planning without human intervention. The single-step retrosynthetic model sets a new state of the art for predicting reactants as well as reagents, solvents and catalysts…
Single-step samplers generate high-quality samples efficiently.
Adversarial examples are malicious inputs designed to fool machine learning models. They often transfer from one model to another, allowing attackers to mount black box attacks without knowledge of the target model's parameters. Adversarial training is the process of explicitly training a model on adversarial examples,…
Lazy neural networks are vulnerable to adversarial attacks.
A novel optimisation framework through quadratic nonlinear projection is introduced for credit portfolio when the portfolio risk is measured by Conditional Value-at-Risk (CVaR). The whole optimisation procedure to search toward the optimal portfolio state is conducted by a series of single-step optimisations under the …
Improved chemical reaction prediction using augmented NLP models.
New method improves adversarial training efficiency and robustness.
CARV reduces compute cost for downstream pipelines using diffusion models.
This paper establishes a theoretical foundation for consistency training in diffusion models.
Unified model improves sampling speed and quality.
The problem of automatic and accurate forecasting of time-series data has always been an interesting challenge for the machine learning and forecasting community. A majority of the real-world time-series problems have non-stationary characteristics that make the understanding of trend and seasonality difficult. Our int…
Efficient algorithm for learning from indirect feedback in complex decision-making scenarios.
A new method reduces compounding errors in model-based reinforcement learning.
Certified training improves robustness against adversarial attacks.
Adversarial examples are perturbed inputs designed to fool machine learning models. Adversarial training injects such examples into training data to increase robustness. To scale this technique to large datasets, perturbations are crafted using fast single-step methods that maximize a linear approximation of the model'…
Gradient descent memorizes many Gaussians efficiently.
We show that the lack of arbitrage in a model with both fixed and proportional transaction costs is equivalent to the existence of a family of absolutely continuous single-step probability measures, together with an adapted process with values between the bid-ask spreads that satisfies the martingale property with resp…
A new method improves stochastic gradient descent for faster and more efficient estimation.
We study an exploration method for model-free RL that generalizes the counter-based exploration bonus methods and takes into account long term exploratory value of actions rather than a single step look-ahead. We propose a model-free RL method that modifies Delayed Q-learning and utilizes the long-term exploration bonu…
In this work, we describe practical lessons we have learned from successfully using contextual bandits (CBs) to improve key business metrics of the Microsoft Virtual Agent for customer support. While our current use cases focus on single step einforcement learning (RL) and mostly in the domain of natural language proce…
When environmental interaction is expensive, model-based reinforcement learning offers a solution by planning ahead and avoiding costly mistakes. Model-based agents typically learn a single-step transition model. In this paper, we propose a multi-step model that predicts the outcome of an action sequence with variable …
Model predicts stock price changes and forecasts using tokenized data.
Closed-form pricing method for multi-asset options.
Many applied decision-making problems have a dynamic component: The policymaker needs not only to choose whom to treat, but also when to start which treatment. For example, a medical doctor may choose between postponing treatment (watchful waiting) and prescribing one of several available treatments during the many vis…
We have developed a novel prediction method based on string invariants. The method does not require learning but a small set of parameters must be set to achieve optimal performance. We have implemented an evolutionary algorithm for the parametric optimization. We have tested the performance of the method on artificial…
Single gradient step finds adversarial examples in random neural networks.
New Calabi-Yau metrics found on complex symmetric spaces.
To act and plan in complex environments, we posit that agents should have a mental simulator of the world with three characteristics: (a) it should build an abstract state representing the condition of the world; (b) it should form a belief which represents uncertainty on the world; (c) it should go beyond simple step-…
A new approach RA improves stochastic optimization by executing multiple steps between subsample updates.
In this paper we study decomposition methods based on separable approximations for minimizing the augmented Lagrangian. In particular, we study and compare the Diagonal Quadratic Approximation Method (DQAM) of Mulvey and Ruszczyński and the Parallel Coordinate Descent Method (PCDM) of Richtárik and Takáč. We show that …
Over the last few years, there has been growing interest in learning models for physically grounded language understanding tasks, such as the popular blocks world domain. These works typically view this problem as a single-step process, in which a human operator gives an instruction and an automated agent is evaluated …
Geodesic descent optimizes likelihood in dually flat spaces.
Density estimation is a fundamental problem in statistical learning. This problem is especially challenging for complex high-dimensional data due to the curse of dimensionality. A promising solution to this problem is given here in an inference-free hierarchical framework that is built on score matching. We revisit the…
Unified finetuning of all quantization degrees of freedom achieves state-of-the-art 4-bit quantization.
There are a multitude of methods to perform multi-set correlated component analysis (MCCA), including some that require iterative solutions. The methods differ on the criterion they optimize and the constraints placed on the solutions. This note focuses perhaps on the simplest version, which can be solved in a single s…
Proposes SD-KDE for density estimation using debiased kernel density with score-based adjustments.
Proposes method for eliciting non-parametric joint priors using normalizing flows.
This work investigates a mixture of LMC and RMHMC with MMALA for geometric ergodicity.
M-FFF generates data on manifolds with fast sampling.
TKFT models computation via smooth vector fields, simulating functions in a single dynamical step.
Deep neural networks excel at learning the training data, but often provide incorrect and confident predictions when evaluated on slightly different test examples. This includes distribution shifts, outliers, and adversarial examples. To address these issues, we propose Manifold Mixup, a simple regularizer that encoura…
New algorithm solves complex equations using deep learning.
Hopfield networks improve reaction template prediction for few/zero-shot scenarios.
PaGoDA reduces diffusion model training costs by 64x.