Spectral measurements reveal hidden representation geometry in language model training.
problem Hidden internal representation in language model training is hard to examine.
method Empirical protocol using activation covariance and per-sample gradient SVD spectra.
result Batch size affects representation geometry, and activation spectra predict token efficiency.
Language models fail to process hallucinated responses, and this study diagnoses the failure.
problem Language models fail to process hallucinated responses, leading to over-concentration or diffuse attention.
method The study uses forced scoring of benchmark-labeled responses to compute attention shapes and analyze the symmetric component of the degree-normalized attention operator.
result The study proves that every transpose-invariant spectral diagnostic of the attention operator is orientation-blind and bounds the sensitivity of any Lipschitz diagnostic by the asymmetry coefficient \(G\).
Persistence diagrams from random matrices follow RMT universality, offering a new spectral diagnostic.
problem Understanding spectral properties of random matrices using topological data analysis.
method Applying Morse theory to persistence diagrams of quadratic forms restricted to unit spheres.
result Persistence entropy outperforms traditional level spacing ratios in discriminating random matrix ensembles.
Model tracks structural changes in Brownian particle configurations on a sphere.
problem Tracking structural changes in Brownian particle configurations on a sphere.
method Introduces Frustrated Distance Matrix (FDM) model for dynamic distance matrices on S^2.
result Preserves static BBS template with dynamics as redistributed spectral mass.
Unified framework detects overfitting in crash classification models.
problem Evaluation metrics fail to detect overfitting in crash classification models.
method Random Matrix Theory and Heavy-Tailed Self-Regularization framework applied to various model types.
result Power-law exponent α reliably distinguishes well-regularized from overfit models.
New framework uses geometry of embeddings to predict robustness.
problem Monitoring robustness in models without OOD labels.
method Constructs graphs from embeddings, measures spectral complexity and curvature.
result Representation geometry predicts robustness reliably.
Nonlinear dynamic volatility has been observed in many financial time series. The recently proposed quantile periodogram offers an alternative way to examine this phenomena in the frequency domain. The quantile periodogram is constructed from trigonometric quantile regression of time series data at different frequencie…
We identify spectral conditions for reliable neural probe interpretation.
problem Unreliable performance of linear probes in interpreting neural representations.
method Formalized Spectral Identifiability Principle (SIP) based on eigengap and Fisher error.
result Reliability of neural probes depends on the eigengap relative to Fisher estimation error.
Spectral Adaptive Conformal Prediction for Structured Non-Exchangeable Data
problem Improving prediction intervals for non-exchangeable time-indexed datasets
method Spectral adaptive conformal prediction
result Improves on fixed spectral weighting while monitoring uncertainty changes
Graph signal processing detects hallucinations in large language models.
problem Detecting factual reasoning from hallucinations in large language models.
method Modeling transformer layers as dynamic graphs, using spectral analysis to define diagnostics.
result Spectral signatures can distinguish different types of hallucinations and achieve high accuracy.
New framework detects directional influence in multivariate time series.
problem Detecting directional influence in multivariate time series.
method Order-constrained spectral non-invariance.
result Unique diagnostic functional for directional influence.
Spectral portfolio theory links neural networks to wealth dynamics via SGD weight matrices.
problem Understanding wealth dynamics from neural network training.
method Direct identification of weight matrices as portfolio allocation matrices, linking SGD forces to portfolio dynamics.
result Spectral properties of SGD weight matrices transition between additive and multiplicative regimes, influencing wealth dynamics.
Bayesian analysis reveals epistemic uncertainty as a key diagnostic for delayed generalization in in-context learning.
problem Delayed generalization in in-context learning from few examples.
method Bayesian perspective, modular arithmetic tasks, approximate Bayesian techniques, spectral mechanism analysis.
result Epistemic uncertainty collapses sharply when the model groks, indicating a practical diagnostic of generalization.
The paper introduces a diagnostic method to detect grokking transitions in models before test accuracy improves.
problem Detecting the transition from training to generalization in machine learning models.
method Summarize task-dependent observables as empirical distributions, map them to Wasserstein/quantile coordinates, and analyze using Hankel dynamic mode decomposition.
result The diagnostic method achieves AUROC \(\approx\) 0.93 for grokking-vs-non-grokking discrimination at the run level.
A registration-free framework monitors shape and color in 4D point clouds.
problem Monitoring shape and color changes in complex parts without registration.
method Laplace-Beltrami operator spectral properties for geometric and color feature capture; combined monitoring scheme for shape and color anomalies.
result Effective detection of shape deformations and color anomalies without registration or mesh reconstruction.
QNA uses quantum-inspired density operators to diagnose market dependence and structural risk.
problem Lack of unified operator representation for market dependence and structural risk diagnostics.
method Quantum Network of Assets (QNA) framework using density operators.
result QNA entropy remains strongly related to covariance spectral entropy but becomes distinct with multi-feature rolling trajectories.
Electroencephalographic (EEG) monitoring of neural activity is widely used for sleep disorder diagnostics and research. The standard of care is to manually classify 30-second epochs of EEG time-domain traces into 5 discrete sleep stages. Unfortunately, this scoring process is subjective and time-consuming, and the defi…
Optimizer choice affects neural scaling laws, changing the exponent α.
problem The exponent α in neural scaling laws L(N)∝N−α varies with the optimizer used. method Controlled random-feature regression experiments with five optimizer variants and six spectral conditions.
result Preconditioned optimizers yield steeper scaling (larger α), with the α-shift increasing across most of the tested spectral range. Transformer models show distinct spectral fingerprints under voice changes.
problem Detecting architectural biases in transformer models.
method Spectral analysis of attention-induced token graphs.
result Clear architectural signatures in model fingerprints correlate with language-specific behavior.
Detects corruption in agentic models during execution.
problem Inconsistent context, retrieval errors, or adversarial inputs corrupt intermediate steps of reasoning chains.
method Analyzes token graphs induced by attention and computes spectral statistics to emit accept/reject signals.
result A single threshold on the high frequency energy ratio optimally detects context inconsistency in agentic models.
High-dimensional models become unstable when sample size falls below a critical level, leading to a phase transition.
problem Instability in high-dimensional learning models when sample size is insufficient.
method Proved the necessity of a Fisher eigenvalue threshold for stability, introduced Fisher floor for verification.
result A sharp phase transition between reliable concentration and inevitable failure in high-dimensional learning.
Applying machine learning in the health care domain has shown promising results in recent years. Interpretable outputs from learning algorithms are desirable for decision making by health care personnel. In this work, we explore the possibility of utilizing causal relationships to refine diagnostic prediction. We focus…
Establishes statistical and computational bounds for influence diagnostics.
problem Identifying influential datapoints or subsets in machine learning models.
method Finite-sample statistical bounds and computational complexity for influence functions and approximate maximum influence perturbations.
result Established statistical and computational guarantees for influence diagnostics.
The article explains how to estimate confusion matrices for classifiers using unlabeled data.
problem Estimating sensitivity and specificity of binary medical diagnostic tests without gold standard tests.
method Modifying diagnostic test solutions to estimate confusion matrices for classifiers on unlabeled data.
result The approach can be used to estimate accuracy statistics for supervised or unsupervised binary classifiers on unlabeled data.
The paper proves geometric and spectral alignment for deep neural networks.
problem Understanding the singular spectra of deep neural network layers.
method Proves deterministic quotient-geometric estimates for singular spectra of Frobenius-normalized layer factors.
result Exact power-law spectra form a trace-normalized Cartan orbit under Frobenius normalization.
Characterizes Hessian eigenspectra for realistic nonlinear models.
problem Understanding Hessian eigenspectra in realistic nonlinear models.
method Deterministic equivalent techniques from random matrix theory.
result Hessian can have qualitatively different spectral behaviors.
AI enhances cancer diagnostics using spectroscopy.
problem Early and accurate cancer diagnosis.
method Combining AI with spectroscopy-based techniques.
result AI improves cancer diagnostics speed and safety.
Paper introduces a diagnostic for approximate inference methods.
problem Estimating errors in probabilistic inference algorithms, especially for approximate methods.
method Repeatedly simulate datasets from the prior and perform inference on each, estimating a symmetric KL-divergence.
result A diagnostic for approximate inference methods can be estimated using symmetric KL-divergence.
Machine learning improves diagnostic test accuracy for bovine tuberculosis.
problem Improving diagnostic test sensitivity for bovine tuberculosis.
method Machine learning to assess risk landscapes and predict infection.
result Test sensitivity improved, detecting 240 more infected herds per year.
AI predicts medical specialty diagnostic choices from EHR records.
problem Predicting timely medical specialty diagnostic workups for patients.
method Ensemble of feed-forward neural networks trained on EHR data.
result Significantly higher accuracy compared to traditional checklists.
The paper calibrates shrinkage covariance estimators for spectral functionals in high dimensions.
problem Calibrating shrinkage covariance estimators for spectral functionals in high dimensions.
method Derives first-order null laws, distribution-free Davis-Kahan bands, and calibrated tests for spectral functionals under shrinkage.
result Calibrated tests and intervals for spectral functionals are provided, addressing the issue of estimation noise and shrinkage bias.
New model addresses instability in hierarchical clustering of social networks.
problem Instability in existing hierarchical clustering algorithms for social networks.
method Introduce T-Stochastic Graphs, a probabilistic model that relaxes ultrametric assumptions. result Prove spectral approach combining Neighbor-Joining is statistically consistent.
VDA improves disentanglement of latent representations in complex signals.
problem Learning disentangled and interpretable representations in nonstationary, high-dimensional time-evolving signals.
method Variational decomposition autoencoding (VDA) framework, incorporating signal decomposition, contrastive self-supervised task, and variational prior approximation.
result DecVAEs surpass state-of-the-art VAE-based methods in disentanglement quality and generalization.
Graph-based approach repairs programs from diagnostic feedback.
problem Learning to repair programs from limited labeled data and compiler error messages.
method Introduces program-feedback graph and graph neural network for reasoning, and self-supervised learning with unlabeled programs.
result DrRepair significantly outperforms prior work, achieving high repair rates.
The paper analyzes convergence in SGD with momentum and proposes a diagnostic test.
problem Detecting convergence in stochastic gradient descent with momentum.
method Analyzes the transient and stationary phases of SGD with momentum, constructs a statistical diagnostic test.
result The proposed diagnostic test effectively detects convergence in the stationary phase of SGD with momentum.
Gradient descent near stability threshold shows sharpness oscillations.
problem Understanding sharpness and stability in non-Euclidean norms during gradient descent.
method Interpreted EoS through Directional Smoothness, defined generalized sharpness for arbitrary norms.
result Non-Euclidean GD exhibits sharpness oscillations around the stability threshold.
Paper proposes a coupling-based diagnostic for SGD stepsize optimization.
problem Optimizing stepsize for SGD convergence.
method Coupling-based convergence diagnostic for monitoring stationarity.
result Proposed stepsize scheme achieves superior performance across convex and non-convex problems.
Gradient descent near stability threshold exhibits sharpness oscillations.
problem Understanding sharpness behavior near stability threshold in non-Euclidean norms.
method Interpreted EoS through Directional Smoothness and generalized sharpness under arbitrary norms.
result Non-Euclidean GD with generalized sharpness shows sharpness oscillations near 2/η. Paper improves predictive distributions for rare events using a simple framework.
problem Local miscalibration of predictive distributions for rare events.
method Semiparametric diagnostic transport maps to correct tail probabilities.
result Semiparametric maps improve predictions for severe weather hazards.
TADDAA improves accuracy diagnostics for variational approximations.
problem Challenges in evaluating the accuracy of variational approximations.
method Uses many short parallel MCMC chains to obtain lower bounds on the error of each posterior functional of interest.
result Validates the practical utility and computational efficiency of TADDAA on various models.
Proposes a diagnostic method to evaluate factor models using cap-axis integrals.
problem Improving factor model evaluation in low-dimensional spaces.
method Lifts pricing errors into a bridge-alpha curve along the market-capitalization rank axis.
result The cap-axis norm is distinct from Sharpe gain and size exposure.
Proposes a diagnostic method to evaluate factor models using cap-axis integrals.
problem Improving factor model evaluation for low-dimensional models.
method Lifts pricing errors into a bridge-alpha curve along the market-capitalization rank axis.
result The cap-axis norm is distinct from Sharpe gain and size exposure.
The receiver operating characteristic (ROC) curve is a very useful tool for analyzing the diagnostic/classification power of instruments/classification schemes as long as a binary-scale gold standard is available. When the gold standard is continuous and there is no confirmative threshold, ROC curve becomes less useful…
The high-dimensionality and volume of large scale multistream data has inhibited significant research progress in developing an integrated monitoring and diagnostics (M&D) approach. This data, also categorized as big data, is becoming common in manufacturing plants. In this paper, we propose an integrated M\&D approach…
Recent studies have shown promising results in using Deep Learning to detect malignancy in whole slide imaging. However, they were limited to just predicting positive or negative finding for a specific neoplasm. We attempted to use Deep Learning with a convolutional neural network algorithm to build a lymphoma diagnost…
Interpretable deep learning is a fundamental building block towards safer AI, especially when the deployment possibilities of deep learning-based computer-aided medical diagnostic systems are so eminent. However, without a computational formulation of black-box interpretation, general interpretability research rely hea…
Online symptom checkers have significant potential to improve patient care, however their reliability and accuracy remain variable. We hypothesised that an artificial intelligence (AI) powered triage and diagnostic system would compare favourably with human doctors with respect to triage and diagnostic accuracy. We per…
Reducing ICD-10 code granularity improves cost model accuracy and stability.
problem High-dimensional regression with ICD-10 codes leads to unstable coefficient estimates.
method Log-linear analytics approach to cost model regularization through diagnostic code merging.
result Reducing ICD-10 code granularity from 7 characters to 6 or fewer improves model interpretability and consistency.