Library · Machine learning & AI
What Probabilistic Latent Modeling got wrong, from 55 theses
Summary written by a model from the records below: The records document frequent failures and design rejections across probabilistic latent modeling techniques including topic models, variational autoencoders, Gaussian processes, and factor analyses. Practitioners regularly encountered issues such as severe sparsity in short texts, inappropriate prior distributions, optimization instabilities, and underperformance relative to simpler deterministic or baseline models. These records come from graduate theses, 2021 to 2026, each labeled with its degree where the repository states it, and each links to its thesis. They were extracted by language models reading the full text, so treat each as a lead to read, not a verdict.
Plus 9 problems the authors left open, listed at the end.
Failure modes, by theses reporting them
Select a mode to jump to its records.
Latent Dirichlet Allocation fails on short and sparse text corpora
Standard Latent Dirichlet Allocation struggles with document sparsity and insufficient word co-occurrences when applied to tweets, short surveys, search queries, and brief text spans. These sparse inputs cause topic incoherence, misalignment with human semantic judgments, and inferior performance compared to alternatives like Dirichlet multinomial mixtures or neural embeddings.
Tried and failed
Latent Dirichlet Allocation topic modelling applied to short-text survey responses. Outcome: worse than baseline. Reason: standard LDA struggles with sparsity and lack of word co-occurrence in short texts
Examining structural, systemic, and enabling approaches to sustainable food system transformations · Imperial · PhD
Tried and failed
Latent Dirichlet Allocation topic modeling applied to short social media texts. Outcome: worse than baseline. Reason: sparse word co-occurrences in short documents led to poor topic coherence compared to NMF and neural embeddings
Assessing mental wellbeing in urban areas using social media data: understanding when and where urbanites stress and de-stress · Georgia Tech · PhD
Considered and rejected
Considered and rejected: Rejected Latent Dirichlet Allocation (LDA) text topic features due to poor perplexity/interpretability on 89 profile tweet aggregations.
Personality-Driven Social Media Curation: How Personality Traits Affect Following Decisions on Twitter · University of Oregon · PhD
Considered and rejected
Considered and rejected: Rejected Latent Dirichlet Allocation (LDA) for short sanction text messages, choosing biterm topic modeling instead.
Three Papers on Peer Sanctioning, its Evaluation, and its Justification · Duke · PhD
Tried and failed
Latent Dirichlet Allocation topic modeling applied to short user comment text corpus. Outcome: no signal. Reason: failed to reveal distinct latent themes or nuanced categorization across user comments
Public Use of Open Access Research: Evidence from the National Academies and Harvard DASH Repository · Georgia Tech · PhD
Tried and failed
Latent Dirichlet Allocation topic modeling applied to short conversational text queries. Outcome: no signal. Reason: text length was too brief and sparse to extract coherent topics
Evaluating the explainability of AI-driven clinical decision support tools · Imperial · PhD
Considered and rejected
Considered and rejected: Rejected Latent Dirichlet Allocation (LDA) in favor of the Gibbs Sampling Dirichlet Multinomial Mixture (GSDMM) model for topic modeling due to GSDMM's superior handling of short, dense tweet text
Decision support for sustainable energy systems, energy economics, urban mobility, and emerging methods in information systems research · Leibniz Universität Hannover · PhD
Considered and rejected
Considered and rejected: Rejected Latent Dirichlet Allocation (LDA) for short survey priority responses due to poor performance on documents with fewer than 50 words.
Examining structural, systemic, and enabling approaches to sustainable food system transformations · Imperial · PhD
Considered and rejected
Considered and rejected: Rejected standard Latent Dirichlet Allocation (LDA) for clustering short extracted action spans due to severe sparsity and inability of synonyms to co-occur in 1-10 word spans.
Considered and rejected
Considered and rejected: Rejected Latent Dirichlet Allocation (LDA) for query topic modeling because short search queries (avg 5.2 words) yield sparse vectors and ignore word sequence order.
EFFECTS OF INFORMATION ON DIGITAL PLATFORMS AND ONLINE ECOSYSTEMS · Penn · PhD
Continuous Gaussian latent priors in variational autoencoders degrade clustering and domain representations
Imposing standard Gaussian latent priors forces an overlapping or overly dense latent space that obscures distinct cluster boundaries and underperforms deterministic autoencoders. These prior assumptions also fail to capture non-Gaussian data statistics, suppress high-frequency dynamics in physical systems, and create latent gaps that harm downstream performance.
Lost to a baseline
Vanilla autoencoders achieved slightly higher silhouette scores for cluster separation compared to VAE due to VAE's continuous, overlapping latent space.
Machine Understanding of Architectural Space: From Analytical to Generative Applications · EPFL · PhD
Considered and rejected
Considered and rejected: Rejected stochastic latent sampling in VAEs, choosing deterministic autoencoders with explicit decoder regularization and ex-post density estimation.
Reining in the Deep Generative Models · Tübingen · PhD
Considered and rejected
Considered and rejected: Rejected standard VAE in favor of deterministic autoencoder because generative latent distribution modeling was unnecessary for embedding compression.
Neural Compression for Scalable Question-Answer Retrieval · Dalhousie · Master's
Considered and rejected
Considered and rejected: Rejected enforcing a Gaussian prior via adversarial discriminators on the style latent bottleneck due to training instability and degraded RMSE
Using wearable sensors and machine learning to enrich lower limb assistive devices · University of Washington · PhD
Tried and failed
variational autoencoders with Gaussian priors applied to systems with strongly non-Gaussian statistics. Outcome: did not generalise. Reason: standard Gaussian latent priors poorly represent strongly non-Gaussian statistics
Physics-Driven Machine Learning for Applications in Geophysical Fluid Dynamics · MIT · PhD
Tried and failed
variational autoencoder for dimensionality reduction before clustering applied to high-dimensional mass spectrometry imaging data. Reason: Gaussian prior forced an overly dense latent space, preventing separation into distinct clusters
Mass spectral imaging of clinical samples using deep learning · Imperial · PhD
Tried and failed
variational autoencoder transformer reduced order model applied to turbulent fluid flow prediction. Reason: Smooth latent representations underpredicted kinetic energy and lost high-frequency spectral content
A Dynamic Model of Unpiloted Aerial Systems for Complex Ship Airwake Environments · Georgia Tech · PhD
Lost to a baseline
Variational Autoencoders (VAEs) lose to standard Autoencoders (AEs) when training SNR and testing SNR perfectly match due to gaps in the latent space
Model and data driven approaches to wireless image transmission · Imperial · Doctorate
Considered and rejected
Considered and rejected: Rejected Variational Autoencoders (VAEs) for sound-masking noise generation because GANs avoid deterministic bias and perform better with discrete latent variables.
Robust Defenses Against Adversarial Machine Learning in Internet of Things Security · Carleton University · PhD
Considered and rejected
Considered and rejected: Rejected Variational Autoencoders (VAEs) in FASTER-CE to avoid assuming Gaussian distributions over tabular latent spaces
Multi-objective approaches towards trustworthy machine learning · UT Austin · PhD
Variational autoencoders suffer from posterior collapse, sampling failures, and latent regularization instabilities
Attempts to enforce latent sparsity or structured covariances in variational autoencoders often fail to prevent posterior collapse or maintain consistent regularization. In addition, loose lower bounds, training data sparsity, and decoupled optimization schemes cause severe artifacts during latent space sampling and model inversion.
Tried and failed
L1 or L2 regularization on posterior parameters applied to variational autoencoder latent representations. Outcome: worse than baseline. Reason: Failed to increase representation sparsity compared to vanilla variational autoencoder baseline.
Injecting Inductive Biases into Distributed Representations of Text · Cambridge · PhD
Tried and failed
variational autoencoder for sparse latent representation applied to text sentence embeddings. Outcome: unstable. Reason: struggled to achieve steady and consistent sparsity across hyperparameter configurations
Injecting Inductive Biases into Distributed Representations of Text · Cambridge · PhD
Tried and failed
autoregressive latent covariance structure applied to variational autoencoders. Reason: did not prevent posterior collapse when it occurred in standard VAE baseline
Deep Time: Deep Learning Extensions to Time Series Factor Analysis with Applications to Uncertainty Quantification in Economic and Financial Modeling · Virginia Tech · Master's
Tried and failed
i.i.d. latent sampling in variational autoencoders applied to set and graph generation. Reason: It optimizes an excessively loose lower bound on the true evidence lower bound.
Equivariant Neural Architectures for Representing and Generating Graphs · EPFL · PhD
Lost to a baseline
Separately trained predictor and VAE required significantly more optimization steps and exhibited higher variance during latent space search compared to simultaneously trained VAE with predictor
Integrating AI across the Chemistry Discovery Cycle: Advancing Sustainable Chemistry through Digital Methods · EPFL · PhD
Tried and failed
conditional variational autoencoder with factorised latent space applied to 3D multi-object scene completion. Reason: failed to learn a structured prior, yielding incomplete reconstructions when sampling from Gaussian noise
Scene understanding for 3D multi-object scenes: labelling, reasoning and decomposing · Imperial · PhD
Tried and failed
latent space sampling in variational autoencoder applied to random 3D mesh generation. Outcome: data insufficient. Reason: training data sparsity led to low valid generation success rate when sampling the latent space randomly
Towards human-centered generative design : cross-modal synthesis for three-dimensional design concept generation · UT Austin · PhD
Tried and failed
latent space inversion via optimization and encoder applied to deep progressive generative adversarial networks. Outcome: did not converge. Reason: optimization and encoder-based methods fail to accurately invert deep multi-scale generative models
Dissection of Deep Neural Networks · MIT · PhD
Considered and rejected
Considered and rejected: Rejected axis-aligned latent disentanglement (e.g., beta-VAE) for tool synthesis in favor of task-classifier-guided latent space traversal.
Learning physical intuition for robotic manipulation · Oxford · PhD
Linear factor analysis and principal component analysis fail to capture complex latent factors and measurement error
Principal component analysis was repeatedly rejected because it assumes zero measurement error, fails to model latent constructs through covariance, and cannot handle time-series structures. Furthermore, retaining high-order components captured sample noise rather than weak signals, causing out-of-sample degradation and poor dimensionality reduction compared to nonlinear approaches.
Tried and failed
projecting latent factor decompositions to recover unprojected components applied to stochastic discount factor decomposition. Outcome: did not converge. Reason: projected weak-factor and idiosyncratic components failed to converge to their true unprojected counterparts
Asset pricing with unsystematic risk · Imperial · PhD
Tried and failed
retaining high-order principal components for weak signals applied to linear factor asset pricing models. Outcome: did not generalise. Reason: high-order components captured primarily sample-specific noise rather than true weak latent factors, causing severe out-of-sample degradation
Asset pricing with unsystematic risk · Imperial · PhD
Lost to a baseline
Linear PCA dimensionality reduction on agent-based stochastic data found a latent space size of n - 1 (25x larger than non-linear DR), resulting in poor reducibility compared to non-linear methods.
Reduced Order Non-INtrusive (RONIN) Modeling for Strategic Defense Planning · Georgia Tech · PhD
Considered and rejected
Considered and rejected: Rejected non-parametric Principal Component Analysis (PCA) for latent factor extraction due to lack of time-series structure and inability to propagate parameter uncertainty
Parameter Uncertainty, Cashflow Betas, and Earnings Announcement Premia · University of Oregon · Doctorate
Considered and rejected
Considered and rejected: Rejected principal component analysis (PCA) because it cannot model underlying latent factors via covariance
An Exploration of Fear of Sleep and Experiential Avoidance in the Context of PTSD and Insomnia Symptoms · University of Oregon · PhD
Considered and rejected
Considered and rejected: Rejected factor analysis (FA) in favor of PCA for factor score generation to describe empirical dataset variance without imposing latent variable normality assumptions.
Whole-water systems modelling for sustainable catchment management · Imperial · PhD
Considered and rejected
Considered and rejected: Deterministic PCA with arbitrary truncation threshold (rejected because it incorporates undesired noise and lacks probabilistic modeling of model reduction error)
Data-driven Multi-Scale Modeling with Uncertainty Quantification of Damage in Unidirectional and Woven Composites Using Parametrically-Upscaled Continuum Damage Mechanics (PUCDM) Model · Johns Hopkins · PhD
Considered and rejected
Considered and rejected: Rejected Principal Component Analysis (PCA) in favor of Exploratory Factor Analysis (EFA) because PCA does not model latent constructs with conceptual meaning and assumes zero measurement error.
Evaluating the Efficacy of Talent Identification and Development in the National Hockey League Entry Draft · York University · PhD
Latent Dirichlet Allocation yields uninterpretable, unstable, or poorly aligned topic structures
Latent Dirichlet Allocation frequently generates generic or loosely related word groupings that fail to map to theoretical constructs or predefined taxonomies. The inferred topics also suffer from sensitivity to text ordering, amorphous boundaries, and poor stability compared to non-negative matrix factorization or CorEx.
Tried and failed
Latent Dirichlet Allocation topic modeling applied to organizational text corpora. Outcome: no signal. Reason: generated uninterpretable topics with generic words failing to map to distinct theoretical constructs
Embracing community-centered approaches: Institutional logics and organizational responses in the U.S. museum field · Cornell · PhD
Tried and failed
unsupervised Latent Dirichlet Allocation applied to text classification against predefined taxonomy. Reason: unsupervised topic clusters did not correspond one-to-one with predefined target taxonomy categories
Analytics-Enabled Quality and Safety Management Methods for High-Stakes Manufacturing Applications · MIT · PhD
Considered and rejected
Considered and rejected: Rejected using Latent Dirichlet Allocation (LDA) for document topic modeling due to non-accurate and inconsistent inferring results.
Using Machine Learning And Natural Language Processing To Improve Scientific Processes · Penn · PhD
Considered and rejected
Considered and rejected: Rejected Latent Dirichlet Allocation (LDA) for semi-supervised topic modeling because topics were more amorphous and less interpretable than CorEx.
The Hard-To-Measure Aspects Of Firms · Penn · PhD
Considered and rejected
Considered and rejected: Rejected Latent Dirichlet Allocation (LDA) for topic modeling due to reliance on bag-of-words, requirement of prior topic number determination, and ignoring word ordering/semantics in favor of Top2Vec.
Three Essays on the Use of Health Information Technology to Improve Patient Safety: Cases of Technology Design and Participation in Professional-Only Healthcare Forums · University at Buffalo · PhD
Considered and rejected
Considered and rejected: Rejected Latent Dirichlet Allocation (LDA) for issue report topic extraction and review aspect extraction because it produces loosely-related words when vocabulary is large
Machine Learning Driven Guidance for Software Maintenance: Enhancing Code Management and Features · Queen's University · Doctorate
Considered and rejected
Considered and rejected: Rejected Latent Dirichlet Allocation (LDA) in favor of Non-negative Matrix Factorization (NMF) due to NMF's higher topic coherence and model stability.
Essays on central bank communication · UT Austin · PhD
Tried and failed
undirected latent Dirichlet allocation applied to longitudinal corporate annual report text. Outcome: unstable. Reason: critically dependent on initial text sequence and could not reliably track recurring topics across decades
The application of quantitative methods to the study of firm-level adaptation and investment preferences in the energy sector · Imperial · PhD
Variational autoencoder representations underperform simpler deterministic baselines on prediction and reconstruction
Downstream models built on variational autoencoder latent spaces were beaten by simpler baselines including linear PCA, k-nearest neighbors, and direct reconstruction methods. Nonlinear latent embeddings often degraded prediction accuracy, increased predictive uncertainty, or produced disorganized representations despite higher training complexity.
Tried and failed
sequence-to-sequence variational autoencoder applied to sketch to latent space mapping. Outcome: worse than baseline. Reason: failed to show significant improvement over simpler vector KNN regression despite higher training complexity
Tried and failed
surrogate modeling on variational autoencoder latent space applied to property prediction from spatial representations. Outcome: worse than baseline. Reason: nonlinear latent embedding degraded forward predictions and increased uncertainty compared to linear dimensionality reduction
Neural Inverse Microstructure Design with Bayesian Scale-Bridging · Georgia Tech · PhD
Lost to a baseline
VAE + Gaussian process regression on latent variables lost to a simple null baseline (predicting test QY using the most similar sequence in the training set), which achieved matching correlation.
INVESTIGATING THE STRUCTURAL REQUISITES OF PHOTOPHYSICAL PROPERTIES IN FLUORESCENT BIOMOLECULES · Johns Hopkins · PhD
Lost to a baseline
VAE reconstruction MSE underperformed PCA and NMF when using ≥ 5 latent parameters
Characterising & classifying the local population of ultracool dwarfs with Gaia DR2 and EDR3 · Imperial · PhD
Lost to a baseline
AE and VAE Latent Noise Segmentation achieved lower ARI (0.926 and 0.918) than simple noisy reconstruction clustering (0.982 and 0.977) on the Gradient Occlusion dataset.
Differences in Visual Perception in Humans and Deep Neural Networks · EPFL · PhD
Lost to a baseline
Large VAE with latent size 128 (44.3 AUPRC) was beaten by PCA difference vector baseline (67.3 AUPRC) on the Floods dataset with k=3 history.
Intelligent decision making on-board satellites · Oxford · PhD
Lost to a baseline
Model 4 (shallow 4-layer VAE) achieved slightly lower validation reconstruction loss than the selected 10-layer Model 2, but was rejected because its latent feature representations were disorganized with blurred cluster boundaries.
Investigating the Brain States Behind FMRI Temporal Dynamics Using Frame-based Analysis Methods and Variational Autoencoders · Georgia Tech · PhD
Lost to a baseline
SHRED achieved higher reconstruction error than the Convolutional Autoencoder (CAE) on the Kuramoto-Sivashinsky system (at L=22 and L=44) and the Kolmogorov flow at Re=30 across varying latent dimensions.
Data-driven recurrent models to forecast and control high-dimensional dynamical systems · University of Washington · PhD
Gaussian process latent models and surrogates struggle with environmental drift and high-dimensional spaces
Standard Gaussian process surrogates failed to track time-varying environmental shifts and were outperformed by gradient boosting baselines during Bayesian optimization. Practitioners also encountered poor uncertainty quantification from Gaussian posterior approximations and severe hyperparameter complexity when scaling latent Gaussian processes to high dimensions.
Tried and failed
Gaussian approximation of deep Gaussian process posteriors applied to Bayesian optimization uncertainty estimation. Reason: yielded poor uncertainty quantification and inaccurate variance estimates for acquisition functions
Physics-informed Machine Learning for Digital Twins of Metal Additive Manufacturing · Virginia Tech · Doctorate
Tried and failed
Standard Bayesian optimization without contextual modeling applied to optimization under dynamic periodic environmental drift. Outcome: did not converge. Reason: Standard Gaussian process regression could not track or account for dynamic time-varying environmental context shifts
Machine Learning Applications for Improving Accelerator Operations · Cornell · PhD
Lost to a baseline
Gaussian Process surrogate underperformed GradientBoosting during Bayesian optimization, dipping below p=0.05 around iteration 50.
Data-Driven Design of Recycling-Friendly Aluminium Alloys · MIT · PhD
Lost to a baseline
One latent GP framework performed worse (median RMS = 32.9 m s-1) than the baseline FF' model (RMS = 28.5 m s-1) on synchronous synthetic test data.
The epoch of giant planet migration : searching for young planets within the stellar noise · UT Austin · PhD
Considered and rejected
Considered and rejected: Rejected Gaussian process latent variable models (Bayesian approach) for mocap synthesis due to hyperparameter complexity, kernel selection issues, and loss of nuances in high-dimensional space
Predicting lower limb kinematics and kinetics from internal measurement units using deep learning · Imperial · PhD
Left open by the authors
Problems the authors named and did not get to.
Left open
Investigate the causal implications and identification strategies of latent topics extracted from text data. Blocker: Lacks specific causal hypotheses, targets, identification strategies, or experimental frameworks
LEARNING AND INFERENCE IN DIGITAL MARKETS: METHODS AND APPLICATIONS · Cornell · PhD
Left open
Benchmark co-segregation-based Bayesian linear regression feature selection against standard methods like PCA and random forests for microbial phenotypic GWAS. Blocker: None
Probing Genomic, Metabolic, And Phenotypic Evolution In Microbes Using Comparative And Experimental Evolution Data · Georgia Tech · PhD
Left open
Develop a probabilistic model of microbial biomass growth to scale column-level dynamics to ecosystem-level microbial community behavior. Blocker: Lack of mathematical formulation, target datasets, and specific probabilistic framework defined in the thesis
Investigation of microbial community dynamics in soil during variable hydrological forcing · EPFL · PhD
Left open
Extract literature corpora from academic databases beyond Scopus to analyze topic trends via Latent Dirichlet Allocation. Blocker: None
Error component models for traffic crash severity analysis · Texas Tech · PhD
Left open
Incorporate monotonicity, fairness, or probabilistic/Bayesian priors into constrained low-rank approximation frameworks. Blocker: Lacks specific mathematical formulations, optimization objectives, and evaluation datasets or metrics.
Algorithms for Data Fusion, Representation Learning, and Scalable Clustering based on Constrained Low-Rank Approximation · Georgia Tech · PhD
Left open
Initialize variational autoencoders or graph-based multi-scale clustering algorithms with latent factors learned from Integrative Hierarchical Poisson Factorisation. Blocker: None
Robust machine learning methods for high-dimensional datasets with applications in genomics and finance · Imperial · PhD
Left open
Extract high-level semantic features from visual stimuli to model and explain latent aesthetic taste criteria among collaborative filtering user cohorts. Blocker: None
TOWARDS UNDERSTANDING AND PREDICTING INDIVIDUALIZED AESTHETIC RESPONSES TO ECOLOGICALLY VALID VISUAL STIMULI · Cornell · PhD
Left open
Develop unsupervised methods to select subsets of latent dimensions in variational autoencoders for anomaly detection. Blocker: None
Anomaly Detection via Latent Variables Learned by Variational Autoencoders · Queen's University · PhD
Left open
Model agent intent using a continuous latent space with variational autoencoders for approximate inference in multiagent reinforcement learning. Blocker: None
Robust and Scalable Multiagent Reinforcement Learning in Adversarial Scenarios · MIT · PhD
Who has worked on this
Authors of the public theses cited on this page, most-cited first.
- Massimo Dello Preite · Imperial · PhD · 2026
Asset pricing with unsystematic risk - Matt Gibson · Imperial · PhD · 2023
Examining structural, systemic, and enabling approaches to sustainable food system transformations - Victor Prokhorov · Cambridge · PhD · 2021
Injecting Inductive Biases into Distributed Representations of Text - Brenden Michael Oates · Georgia Tech · PhD · 2026
A Dynamic Model of Unpiloted Aerial Systems for Complex Ship Airwake Environments - Bianca Champenois · MIT · PhD · 2026
Physics-Driven Machine Learning for Applications in Geophysical Fluid Dynamics - Luca Francesco Montanelli · MIT · PhD · 2026
Data-Driven Design of Recycling-Friendly Aluminium Alloys - Jan P Williams · University of Washington · PhD · 2026
Data-driven recurrent models to forecast and control high-dimensional dynamical systems - Elif Celikors · Cornell · PhD · 2025
TOWARDS UNDERSTANDING AND PREDICTING INDIVIDUALIZED AESTHETIC RESPONSES TO ECOLOGICALLY VALID VISUAL STIMULI - Dongjin Choi · Georgia Tech · PhD · 2025
Algorithms for Data Fusion, Representation Learning, and Scalable Clustering based on Constrained Low-Rank Approximation - Tobias Kraschewski · Leibniz Universität Hannover · PhD · 2025
Decision support for sustainable energy systems, energy economics, urban mobility, and emerging methods in information systems research - Moamen Khiet · Dalhousie · Master's · 2025
Neural Compression for Scalable Question-Answer Retrieval - Oliver Tobias Schilter · EPFL · PhD · 2025
Integrating AI across the Chemistry Discovery Cycle: Advancing Sustainable Chemistry through Digital Methods
Checking a claim in this area?
We can run the same search on any method or claim. If nothing turns up, we will say so, and that proves nothing on its own.