REVIEW 12 cited by
On the Pitfalls of Heteroscedastic Uncertainty Estimation with Probabilistic Neural Networks
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
On the Pitfalls of Heteroscedastic Uncertainty Estimation with Probabilistic Neural Networks
read the original abstract
Capturing aleatoric uncertainty is a critical part of many machine learning systems. In deep learning, a common approach to this end is to train a neural network to estimate the parameters of a heteroscedastic Gaussian distribution by maximizing the logarithm of the likelihood function under the observed data. In this work, we examine this approach and identify potential hazards associated with the use of log-likelihood in conjunction with gradient-based optimizers. First, we present a synthetic example illustrating how this approach can lead to very poor but stable parameter estimates. Second, we identify the culprit to be the log-likelihood loss, along with certain conditions that exacerbate the issue. Third, we present an alternative formulation, termed $\beta$-NLL, in which each data point's contribution to the loss is weighted by the $\beta$-exponentiated variance estimate. We show that using an appropriate $\beta$ largely mitigates the issue in our illustrative example. Fourth, we evaluate this approach on a range of domains and tasks and show that it achieves considerable improvements and performs more robustly concerning hyperparameters, both in predictive RMSE and log-likelihood criteria.
Forward citations
Cited by 12 Pith papers
-
Walking on Heat Stars for Parabolic Heat Equations with Neumann Boundary Conditions
Walk on Heat Stars provides a boundary-integral Monte Carlo solver for parabolic PDEs with Neumann conditions via exact heat-ball sampling that yields unbiased estimators.
-
From Pixels to Newtons: Predicting In Vivo Joint Contact Forces from Monocular Video
A transformer model predicts in vivo hip and knee contact forces from uncalibrated monocular video at accuracy matching subject-specific musculoskeletal simulations under leave-one-subject-out validation.
-
Variance-aware Reward Modeling with Anchor Guidance
Anchor-guided variance-aware reward modeling uses two response-level anchors to resolve non-identifiability in Gaussian models of pluralistic preferences, yielding provable identification, a joint training objective, ...
-
Bridging Ab Initio Symmetries and Global Nuclear Masses with Interpretable Neural Networks
Symmetry-informed neural networks using SU(3)/SU(4) Casimir operators achieve lower RMSE on global nuclear masses than liquid-drop models, with WINN reaching 0.430 MeV validation error and showing dripline and superhe...
-
A Multi-Agent System for Motor Design Optimization via an FEA-AI Hybrid Approach
A multi-agent LLM framework with retrieval-augmented problem setup, automated FEA data generation, and uncertainty-based FEA-AI switching improves IPMSM iron-loss optimization under a matched simulation budget.
-
Probabilistic denoising for reliable signal extraction in spectroscopy
A probabilistic denoising model recovers spectral features from Poisson-noisy 3D ARPES data at 0.02 electrons per voxel and propagates uncertainties into superconducting gap fits for cuprate superconductors.
-
Monte Carlo PDE Solvers for Nonlinear Radiative Boundary Conditions
A relaxed Picard iteration plus heteroscedastic boundary denoising lets Monte Carlo PDE solvers solve heat equations with nonlinear radiation boundary conditions more accurately than linearization.
-
NeuroSymActive: Differentiable Neural-Symbolic Reasoning with Active Exploration for Knowledge Graph Question Answering
NeuroSymActive combines soft-unification symbolic modules, a neural path evaluator, and Monte-Carlo-style active exploration to reach strong answer accuracy on KGQA benchmarks while cutting graph lookups and model cal...
-
Embedding Linear Equality Constraints in Probabilistic Neural Networks for Dynamic Modelling
Probabilistic neural network framework embeds linear equality constraints for dynamic chemical process modeling, showing improved accuracy, calibration, and constraint adherence on reduced data plus faster training on...
-
A Multi-Agent System for Motor Design Optimization via an FEA-AI Hybrid Approach
A multi-agent system with RAG, FEA, and uncertainty-driven AI surrogates automates IPMSM design optimization and outperforms pure FEA or AI baselines under fixed simulation budgets.
-
NeuroSymActive: Differentiable Neural-Symbolic Reasoning with Active Exploration for Knowledge Graph Question Answering
NeuroSymActive claims state-of-the-art KGQA accuracy (WebQSP 87.1, CWQ 62.5 Hits@1) by coupling differentiable neural-symbolic reasoning with uncertainty-guided MCTS and active human queries.
-
Amplitude Uncertainties Everywhere All at Once
Compares ensemble, Bayesian, and evidential regression approaches for uncertainty quantification in amplitude surrogates and shows they detect localized training data issues.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.