AI Paper Atlas

Research map / Benchmarks / Energy

Energy leaderboard

32 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on Energy. Best RMSE: 0.087 (DCL-NN, NeurIPS 2024).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

Energy · RMSE

13 papers, each with its best reported number for its own method; lower is better.

#RMSE ↓MethodPaperTopic
10.087DCL-NNA Kernel Perspective on Distillation-based Collaborative Learning
NeurIPS 2024 · Table: Performance comparison
Kernels and regression theory
20.31SSM (L2)Supervised Score-Based Modeling by Gradient Boosting
AAAI 2025 · Table: UCI regression RMSE
Variational and generative sampling
30.42WEvidentialWasserstein Gradient Boosting: A Framework for Distribution-Valued Supervised Learning
NeurIPS 2024 · Table: The NLLs and RMSEs
Uncertainty and Gaussian processes
40.46Dα-KPDeep Kernel Posterior Learning under Infinite Variance Prior Weights
ICLR 2025 · UCI out-of-sample errors table
Uncertainty and Gaussian processes
50.461GPNet, 500
RFE
Scalable Training of Inference Networks for Gaussian-Process Models
ICML 2019 · Regression: RMSE table
Uncertainty and Gaussian processes
60.467KIVI
BNN
Kernel Implicit Variational Inference
ICLR 2018 · Table 1
Variational and generative sampling
70.474mBNNMasked Bayesian Neural Networks : Theoretical Guarantee and its Posterior Inference
ICML 2023 · Real dataset table
Uncertainty and Gaussian processes
80.786sgGP (nn)Gaussian Process Parameter Estimation Using Mini-batch Stochastic Gradient Descent: Convergence Guarantees and Empirical Benefits
JMLR 2022 · Table 1
Implicit bias and generalization
90.92GELU kernel GPAvoiding Kernel Fixed Points: Computing with ELU and GELU Infinite Networks
AAAI 2021 · Table: Best performing models
Kernels and regression theory
101.83Ours
DER
The Evidence Contraction Issue in Deep Evidential Regression: Discussion and Solution
AAAI 2024 · Table 1
Prediction and decision losses
3 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Energy · NLL

11 papers, each with its best reported number for its own method; lower is better.

#NLL ↓MethodPaperTopic
10.53WEvidentialWasserstein Gradient Boosting: A Framework for Distribution-Valued Supervised Learning
NeurIPS 2024 · Table: The NLLs and RMSEs
Uncertainty and Gaussian processes
20.6NGBoost
Normal distribution; depth-3 decision trees
NGBoost: Natural Gradient Boosting for Probabilistic Prediction
ICML 2020 · Table 1
Uncertainty and Gaussian processes
30.67mBNNMasked Bayesian Neural Networks : Theoretical Guarantee and its Posterior Inference
ICML 2023 · Real dataset table
Uncertainty and Gaussian processes
41.17MT-ENet
ENet
Improving Evidential Deep Learning via Multi-Task Learning
AAAI 2022 · UCI benchmark table
Uncertainty and Gaussian processes
51.35UR-ERN
ERN
Uncertainty Regularized Evidential Regression
AAAI 2024 · UCI regression results table
Prediction and decision losses
61.36Ours
DER
The Evidence Contraction Issue in Deep Evidential Regression: Discussion and Solution
AAAI 2024 · Table 1
Prediction and decision losses
71.38Deep EnsemblesSimple and Scalable Predictive Uncertainty Estimation using Deep Ensembles
NeurIPS 2017 · Table 1
Uncertainty and Gaussian processes
81.39Evidential regressionDeep Evidential Regression
NeurIPS 2020 · Table 1
Uncertainty and Gaussian processes
92.22V-WTAWinner-takes-all learners are geometry-aware conditional density estimators
ICML 2024 · UCI regression benchmark table
Variational and generative sampling
102.93CDTreeConditional Density Estimation with Histogram Trees
NeurIPS 2024 · Appendix, standard deviation o
Prediction and decision losses
1 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Energy · Test log-likelihood

11 papers, each with its best reported number for its own method; higher is better.

1 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Energy · test RMSE

10 papers, each with its best reported number for its own method; lower is better.

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.