A generative model may generate utter nonsense when it is fit to maximize the likelihood of observed data. This happens due to "model error," i.e., when the true data generating distribution does not fit within the class of generative models being learned. To address this, we propose a model of active distribution lear…
On-device research index
arXiv research
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.
168,811 papers · 148 categories
Trend · papers per month
48 results for “Nonsensical Examples”
Proposes a new method for generating better negative examples in KBC.
problem Random negative sampling generates nonsensical examples that slow down learning and decrease accuracy.
method Distributional Negative Sampling generates meaningful negative examples.
result Significant improvement in Mean Reciprocal Rank values in three benchmarks.
Paper examines NMT robustness to nonsensical inputs.
problem NMT systems fail when source sentences are altered.
method Soft-attention technique to replace words in source sentences.
result Proposed technique achieves high success rate and outperforms existing methods.
Finite-order invariants of knots in arbitrary 3-manifolds (including non-orientable ones) are constructed and studied by methods of the topology of discriminant sets. Obstructions to the integrability of admissible weight systems to well-defined knot invariants are identified as 1-dimensional cohomology classes of gene…
We discuss the local and global problems for the equivalence of geometric structures of an arbitrary order and, in later sections, attention is given to what really matters, namely the equivalence with respect to transformations belonging to a given pseudo-group of transformations. We first give attention to general pr…
Absence-of-Arbitrage (AoA) is the basic assumption underpinning derivatives pricing theory. As part of the OTC derivatives market, the CDS market not only provides a vehicle for participants to hedge and speculate on the default risks of corporate and sovereign entities, it also reveals important market-implied default…
Since machine learning models have been applied to neuroimaging data, researchers have drawn conclusions from the derived weight maps. In particular, weight maps of classifiers between two conditions are often described as a proxy for the underlying signal differences between the conditions. Recent studies have however…
AI models forget statistics' lesson: correlation doesn't imply causation.
problem AI models often produce flawed causal models due to ignoring correlation vs causation.
method Demonstrates examples of flawed AI models and proposes rethinking core models.
result Current efforts to make AI models ethical are insufficient.
Random investment strategies outperform sensible ones, even with forecasts.
problem The usefulness of investment strategies based on forecasts is questioned.
method Investigated the performance of sensible and nonsensical investment strategies, including forecasts.
result There is no substantial difference between the performances of ``best'' and ``trivial'' forecasts.
CNNs overinterpret inputs, leading to high accuracy without meaningful features.
problem High accuracy in image classifiers can mask subtle model failures.
method Batched Gradient SIS method for discovering sufficient input subsets.
result Overinterpretation allows models to make confident predictions with masked input features.
Weakly-supervised RL identifies meaningful tasks, improving performance in complex environments.
problem Learning to efficiently explore and distinguish between meaningful and irrelevant tasks.
method Weak supervision to automatically disentangle meaningful tasks from a large space of nonsensical tasks.
result The learned subspace of meaningful tasks leads to substantial performance gains, especially in complex environments.
Generative Adversarial Networks (GANs) have a great performance in image generation, but they need a large scale of data to train the entire framework, and often result in nonsensical results. We propose a new method referring to conditional GAN, which equipments the latent noise with mixture of Student's t-distributio…
Language models exhibit low-rank structure, which can be used for generation.
problem Understanding the low-dimensional structure of large language models.
method Empirical demonstration and theoretical analysis of the approximate rank of language models' logits.
result Language models can generate responses using linear combinations of unrelated prompts.
Flowing with Confidencestat.ML
Generative models need per-sample confidence scores to improve quality and stability.
problem Generative models produce unreliable outputs and lack confidence measures.
method Flow Matching with Confidence (FMwC) injects noise and integrates it through the network, providing per-sample confidence scores.
result The confidence score correlates with the velocity field's divergence, offering insights into generative processes.
Most successful machine intelligence systems rely on gradient-based learning, which is made possible by backpropagation. Some systems are designed to aid us in interpreting data when explicit goals cannot be provided. These unsupervised systems are commonly trained by backpropagating through a likelihood function. We i…
Multi-Region Neural Representation: A novel model for decoding visual stimuli in human brainsstat.ML
Multivariate Pattern (MVP) classification holds enormous potential for decoding visual stimuli in the human brain by employing task-based fMRI data sets. There is a wide range of challenges in the MVP techniques, i.e. decreasing noise and sparsity, defining effective regions of interest (ROIs), visualizing results, and…
FCNv2 robustness tested under noise and random initial conditions.
problem Assessing AI weather forecasting model robustness to input noise.
method Two experiments with varying noise levels and random initial conditions.
result FCNv2 preserves hurricane features under low to moderate noise, but underestimates intensity and persistence.
Market Microstructure is the investigation of the process and protocols that govern the exchange of assets with the objective of reducing frictions that can impede the transfer. In financial markets, where there is an abundance of recorded information, this translates to the study of the dynamic relationships between o…
Paper introduces SDM for detecting LLM hallucinations, improving on entropy tests.
problem Challenges of Large Language Models (LLMs) with non-factual, nonsensical responses.
method Joint clustering on sentence embeddings to measure semantic divergence between prompts and responses.
result SDM framework detects deeper form of arbitrariness in LLM responses.
Deep neural networks (DNNs) have been widely used in the fields such as natural language processing, computer vision and image recognition. But several studies have been shown that deep neural networks can be easily fooled by artificial examples with some perturbations, which are widely known as adversarial examples. A…
Generates positive examples from noisy data streams.
problem Learning from noisy example streams in hypothesis classes.
method Extending results from previous studies to account for noise.
result Conditions for noisily generatable binary hypothesis classes.
Efficiently computes per-example gradients in CNNs for differential privacy.
problem Computing per-example gradients in CNNs for differential privacy.
method Comparison of existing strategies and introduction of a new per-example gradient calculation method.
result The new method is advantageous depending on model architecture and training.
We construct new examples of self-similar solutions and translating solitons for Lagrangian mean curvature flow by extending the method of Joyce, Lee and Tsui. Those examples include examples in which the Lagrangian angle is arbitrarily small as the examples of Joyce, Lee and Tsui.
With rapid progress and significant successes in a wide spectrum of applications, deep learning is being applied in many safety-critical environments. However, deep neural networks have been recently found vulnerable to well-designed input samples, called adversarial examples. Adversarial examples are imperceptible to …
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.
Cincer cleans both new and past data by identifying and relabeling suspicious and counter-examples.
problem Sequential learning under label noise, especially in applications with human supervision.
method Cincer uses example-based explanations to identify and relabel suspicious and counter-examples, leveraging Fisher information matrix approximation.
result Cincer achieves better data and models by clarifying the model's suspicions, especially with FIM approximation.
Method identifies over-optimized adversarial examples using IQR-based logit thresholding.
problem Over-optimized adversarial examples are hard to detect due to model optimization.
method IQR-based logit thresholding method for identifying over-optimized adversarial examples.
result The logits of the model are more useful in detecting over-optimized adversarial examples.
Gödel's sentence is an adversarial example but unsolvable.
problem The existence and unsolvability of adversarial examples.
method Analysis of non-isomorphic natural explanations and computability theory.
result Gödel's sentence is an adversarial example that cannot be solved.
Enhances adversarial example transferability by fine-tuning existing examples.
problem Adversarial examples are often overfit to a source model, limiting black-box transferability.
method Intermediate Level Attack (ILA) fine-tunes adversarial examples on a pre-specified layer of the source model.
result ILAs achieve high transferability to target models without knowledge of their architecture.
It has been shown that adversaries can craft example inputs to neural networks which are similar to legitimate inputs but have been created to purposely cause the neural network to misclassify the input. These adversarial examples are crafted, for example, by calculating gradients of a carefully defined loss function w…
Linear classifiers can be made robust to strong adversarial examples attacks.
problem Understanding and quantifying adversarial examples in linear classification.
method Proposed a more practical definition of strong adversarial examples, showing robustness to attacks.
result Linear classifiers can be made robust to strong adversarial examples attacks.
Proposes BATer for improved adversarial example detection.
problem Detecting adversarial examples in neural networks.
method Introduces a Bayesian adversarial example detector (BATer) using random components in a Bayesian neural network.
result BATer outperforms state-of-the-art detectors in adversarial example detection.
Catastrophic forgetting of connectionist neural networks is caused by the global sharing of parameters among all training examples. In this study, we analyze parameter sharing under the conditional computation framework where the parameters of a neural network are conditioned on each input example. At one extreme, if e…
Deep neural networks (DNNs) have transformed several artificial intelligence research areas including computer vision, speech recognition, and natural language processing. However, recent studies demonstrated that DNNs are vulnerable to adversarial manipulations at testing time. Specifically, suppose we have a testing …
RelatIF selects more intuitive training examples for explaining model predictions.
problem Influence functions identify outliers as explanatory examples, leading to poor explanations.
method RelatIF separates global and local influence, optimizing for local relative to global effects.
result Examples selected by RelatIF are more intuitive than those from influence functions.
On variational bivectorsmath.DG
New examples of variational bivectors found that are not Poissonian.
problem Identifying variational bivectors that are not Poissonian.
method Constructing examples of variational bivectors.
result Found examples of variational bivectors that are not Poissonian.
Study on adversarial examples and defenses for malware classification.
problem Vulnerability of neural networks to adversarial examples in malware classification.
method Analysis of different approaches for crafting adversarial examples and defense techniques in malware domain.
result Comparison of effectiveness of different approaches on multiple datasets.
We give examples of certain kind of minimal orbits of Hermann actions and discuss whether each of the examples is austere.
Recent studies show that widely used deep neural networks (DNNs) are vulnerable to carefully crafted adversarial examples. Many advanced algorithms have been proposed to generate adversarial examples by leveraging the distance for penalizing perturbations. Researchers have explored different defense met…
Algorithm learns from both labeled and arbitrary test examples, giving guarantees for bounded VC dimension classes.
problem Learning from arbitrary test examples, not just perturbations.
method Selective transductive learning algorithm that outputs abstaining predictions.
result Nontrivial guarantees for bounded VC dimension classes with arbitrary train and test distributions.
When learning a new concept, not all training examples may prove equally useful for training: some may have higher or lower training value than others. The goal of this paper is to bring to the attention of the vision community the following considerations: (1) some examples are better than others for training detector…
Paper defends iris recognition from adversarial examples using wavelet decomposition.
problem Adversarial examples threaten deep neural networks in biometric applications.
method Wavelet domain denoising of input examples to detect and mitigate adversarial attacks.
result Proposed defense strategies improve recognition accuracy against adversarial attacks.
StrokeCoder uses Transformers to generate images from single examples.
problem Creating diverse images from a single example.
method Transformer Neural Network learns from a single path-based example to generate a set of images.
result The model can generate a large set of deviated images that still represent the original image's style and concept.
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 upstream examples after new tasks, using low-rank matrix approximation.
result Forgetting can be predicted efficiently using matrix completion over empirical associations.
Despite the great success achieved in machine learning (ML), adversarial examples have caused concerns with regards to its trustworthiness: A small perturbation of an input results in an arbitrary failure of an otherwise seemingly well-trained ML model. While studies are being conducted to discover the intrinsic proper…
New examples of manifolds with positive scalar curvature and infinitely many poles.
problem Constructing manifolds with positive scalar curvature and understanding their limits.
method Extending previous examples to create new sequences of manifolds.
result Found new examples of manifolds with infinitely many poles and positive scalar curvature.
Paper generates natural adversarial examples for hyperspectral data.
problem Creating adversarial examples for black-box models.
method Modified Wasserstein GAN reweights true data distribution.
result Successfully generates adversarial hyperspectral signatures.
This study explores how examples influence ICL in LLMs.
problem Understanding how examples impact in-context learning in large language models.
method Theoretical study with a probabilistic model extending from Gaussian mixture model.
result The impact of pre-training knowledge and examples on ICL prediction accuracy.