LLMs show surprising confidence in their answers, beyond just tokens.
problem LLMs lack meaningful confidence estimates for their responses.
method Semantic calibration test based on local loss optimality and equivalence classes.
result Base LLMs are semantically calibrated across tasks, contrary to expectations.
The paper examines uncertainty calibration for object detection models in autonomous driving.
problem Uncertainty in object detection predictions and its calibration.
method Definition and evaluation of semantic and spatial uncertainty, calibration methods for uncertainty distributions.
result Calibrated uncertainty improves the overall performance of object detection models in real-world scenarios.
Improved segmentation model adaptation for new domains.
problem Reduced performance of pre-trained models on new domains.
method Calculated soft-label prototypes and predicted closest to class probabilities.
result Significant performance improvements on synthetic-to-real segmentation.
A new framework evaluates LLM calibration in open-ended QA.
problem Evaluating LLM calibration in open-ended QA settings.
method Sem-ECE framework: sampling answers, grouping by semantic classes, and using frequencies as confidence.
result Sem-ECE estimators are unbiased and Sem2 achieves smaller calibration error. This work provides uncertainty intervals for semantic latent variables in disentangled latent spaces.
problem Challenges in providing meaningful uncertainty quantification for semantic information in disentangled latent spaces.
method Uses quantile regression to output heuristic uncertainty intervals, calibrates these intervals to contain true latent values, and propagates them through the generator.
result Reliably communicates semantically meaningful, principled, and instance-adaptive uncertainty in image super-resolution and image completion.
The semantic map calibrates uncertainty from language model probabilities.
problem Uncertainty in language model probabilities for professional decisions.
method Prespecified semantic map linking probabilities of verbal responses to probabilities of declared states.
result Language-derived probabilities outperform printed numerical probabilities and recover valid uncertainty coverage.
New insights show embedding lengths correlate with semantic properties.
problem Contrastive embedding norms ignore embedding magnitudes but correlate with semantic properties.
method Formal theoretical framework and analysis of optimization dynamics.
result Embedding lengths encode semantic information as a byproduct of training.
Improves confidence calibration in neural networks by smoothing labels based on class similarity.
problem Improving confidence calibration in deep neural networks for safety-critical applications.
method Proposes a novel label smoothing technique where label values are based on similarities with the reference class, using different similarity measurements.
result Consistently outperforms state-of-the-art calibration techniques on various datasets and network architectures.
Develops a power-calibrated framework for LLM watermarking, optimizing tradeoffs between detectability and distortion.
problem The trade-off between detectability and semantic distortion in logit-based watermarking.
method Power-calibrated statistical framework for watermark hyperparameters, establishing explicit relationships.
result Derives practical parameter selection procedures achieving optimal tradeoffs under constraints.
A new method for measuring prediction uncertainty in classifiers.
problem Measuring uncertainty of predictions from machine learning methods.
method Density Based Calibration (DBCal) technique.
result Expected calibration error of less than 0.2% on binary classifiers and less than 3% on semantic segmentation networks.
ManifoldMind uses adaptive-curvature probabilistic spheres for trustworthy recommendations in semantic hierarchies.
problem Sparse and abstract recommendation domains where users explore diverse conceptual paths.
method Adaptive-curvature probabilistic spheres, soft multi-hop inference, and curvature-aware semantic kernel.
result Superior NDCG, calibration, and diversity compared to baselines on public benchmarks.
Most distortion correction methods focus on simple forms of distortion, such as radial or linear distortions. These works undistort images either based on measurements in the presence of a calibration grid, or use multiple views to find point correspondences and predict distortion parameters. When possible distortions …
The study examines how much data is needed for generative and vision-language models to make reliable predictions.
problem Ensuring reliable predictions with low data for models used in medical decision support.
method Analyzes uniform convergence bounds for VLM-induced classifiers under low-dimensional semantic representations.
result Finite-sample uniform convergence bounds for accuracy and calibration functionals of VLM-induced classifiers.
Statsformer validates and adapts LLM-derived semantic priors for improved supervised learning.
problem Unreliable semantic priors from LLMs can degrade supervised learning performance.
method Adapts LLM-derived feature scores into a family of learner-specific prior-injection mechanisms, calibrating their influence using out-of-fold validation.
result Improves prediction performance by adaptively downweighting unreliable LLM priors, ensuring a guardrailed statistical learning system.
Enhances uncertainty estimation in medical image segmentation.
problem Frequency-related noise in medical imaging leads to biased uncertainty estimates.
method Extends MC-Dropout to the frequency domain for better uncertainty estimation.
result MC-Frequency Dropout improves calibration and uncertainty in semantic segmentation.
The paper addresses decision making with partially calibrated forecasts, offering a robust approach.
problem Developing a decision-making strategy for forecasts that are only partially calibrated.
method A minimax approach to mapping predictions to actions, considering worst-case distributions.
result The minimax optimal decision rule is to trust predictions and act accordingly, even for partially calibrated forecasts.
Method provides formal guarantees for decomposing model uncertainty.
problem Decomposing model uncertainty into aleatoric and epistemic components.
method Higher-order calibration using k-snapshots.
result Formal guarantees for aleatoric uncertainty matching real-world distribution.
GC-FCP provides efficient federated CP with group-conditional coverage guarantees.
problem Uncertainty quantification in federated settings with distributed calibration data.
method Group-conditional federated conformal prediction (GC-FCP) using group-stratified coresets.
result GC-FCP offers efficient aggregation and calibration compared to centralized methods.
MADOD meta-learns invariant features for OOD detection across unseen domains.
problem Simultaneous covariate and semantic shifts in real-world machine learning applications.
method Meta-learning and G-invariance to learn robust, domain-invariant features.
result Superior performance in semantic OOD detection across unseen domains.
New framework assesses LM uncertainty without thresholding.
problem Uncertainty quantification for LMs, especially comparing different measures.
method Rank-Calibration framework to assess uncertainty and confidence measures.
result Higher uncertainty correlates with lower generation quality.
We introduce generative adversarial models in which the discriminator is replaced by a calibrated (non-differentiable) classifier repeatedly enhanced by domain relevant features. The role of the classifier is to prove that the actual and generated data differ over a controlled semantic space. We demonstrate that such m…
This work tackles uncertainty quantification in language models, proposing a principled approach.
problem Challenges in identifying task-specific uncertainties in large language models.
method Bayesian decision theory, focusing on a similarity measure between generated and hypothetical true responses.
result Derives a measure for epistemic uncertainty based on a missing data perspective.
We propose a calibrated multivariate regression method named CMR for fitting high dimensional multivariate regression models. Compared with existing methods, CMR calibrates regularization for each regression task with respect to its noise level so that it simultaneously attains improved finite-sample performance and tu…
Current approaches in approximate inference for Bayesian neural networks minimise the Kullback-Leibler divergence to approximate the true posterior over the weights. However, this approximation is without knowledge of the final application, and therefore cannot guarantee optimal predictions for a given task. To make mo…
Improved OOD detection across various shifts using multi-encoder fusion of RDMs.
problem Out-of-distribution detection across multiple types of distribution shifts.
method Statistical identification of encoder sensitivity, EncMin2L fusion, and Tippett minimum combination.
result Achieves AUROC ≥ 0.94 across four shift types, outperforming state-of-the-art detectors.
Variance-Calibrated Modulation (VCM) addresses the likelihood trap in LLMs by reshaping the probability distribution before truncation.
problem LLMs fall into the likelihood trap, leading to repetitive degeneration and vocabulary dullness.
method VCM reshapes the probability distribution before truncation through Contextual Searchlight and Adaptive Self-Debiasing.
result VCM mitigates the likelihood trap across open-ended generation, factual QA, and mathematical reasoning.
ECLIPSE detects AI hallucinations in finance with high accuracy.
problem Hallucinations in AI-generated answers limit safe deployment in finance.
method Combines entropy estimation and perplexity decomposition to measure model evidence use.
result ECLIPSE achieves ROC AUC of 0.89 and average precision of 0.90 on financial QA dataset.
The paper presents a method for generating well-calibrated prediction intervals using quality-driven deep ensembles.
problem Generating reliable prediction intervals for regression analysis.
method A multi-objective loss function combining quality measures for prediction intervals and point estimates, with a penalty function to ensure semantic integrity and stability.
result The method produces well-calibrated prediction intervals and point estimates, capturing both aleatoric and epistemic uncertainty.
Language is increasingly being used to define rich visual recognition problems with supporting image collections sourced from the web. Structured prediction models are used in these tasks to take advantage of correlations between co-occurring labels and visual input but risk inadvertently encoding social biases found i…
The paper introduces a method to improve adversarial robustness in neural networks using randomized perturbations.
problem Deep neural networks are sensitive to small perturbations on correctly classified examples, leading to erroneous predictions.
method The approach uses randomized perturbations to optimize the worst case loss function over all possible substitutions of training examples, ensuring that substitution likelihood is weighted by the proximity to the original word.
result The method achieves performance gains and differentially-private model training, improving robustness against adversarial attacks.
Semantic TrueLearn uses semantic graphs to improve educational recommendation systems.
problem Challenges in handling semantic and hierarchical structure in knowledge areas.
method Introduces a novel learner model that exploits semantic relatedness between knowledge components using a Wikipedia link graph.
result Achieves statistically significant improvements in predictive performance for educational engagement.
A framework isolates VQA reasoning from perception for better model evaluation.
problem Improper separation of visual perception and reasoning in VQA models.
method Introducing a framework and a top-down calibration technique to decouple reasoning from perception.
result Improved evaluation of VQA models by separating reasoning from perception.
Transfer learning aims at building robust prediction models by transferring knowledge gained from one problem to another. In the semantic Web, learning tasks are enhanced with semantic representations. We exploit their semantics to augment transfer learning by dealing with when to transfer with semantic measurements an…
This paper presents a Semantic Attribute Modulation (SAM) for language modeling and style variation. The semantic attribute modulation includes various document attributes, such as titles, authors, and document categories. We consider two types of attributes, (title attributes and category attributes), and a flexible a…
The paper shows how integrating categorical semantics can enhance unsupervised domain translation.
problem Improving unsupervised domain translation between perceptually different domains.
method Learning invariant categorical semantic features in an unsupervised manner and conditioning them on the style encoder.
result Conditioning the style encoder on learned categorical semantics improves translation and stylization.
Identifier names convey useful information about the intended semantics of code. Name-based program analyses use this information, e.g., to detect bugs, to predict types, and to improve the readability of code. At the core of name-based analyses are semantic representations of identifiers, e.g., in the form of learned …
New system preserves message meaning in wireless networks, improving data rate.
problem Efficiently transmitting message meaning in wireless networks.
method Modeling semantics as hidden random variables, using Information Bottleneck for compression.
result 20 dB SNR improvement for semantic communication.
New diffusion models improve counterfactual image generation with semantic control.
problem Challenges in preserving identity, maintaining quality, and ensuring causal model faithfulness in counterfactual image generation.
method Integrates semantic representations into diffusion models through Pearlian causality, introducing spatial, semantic, and dynamic abduction.
result Demonstrates high-level semantic identity preservation and principled trade-offs between faithful causal control and identity preservation.
New approach uses SPG for semantic communication without a known channel model.
problem Designing efficient semantic communication systems without a known channel model.
method Applying Stochastic Policy Gradient (SPG) for reinforcement learning.
result Achieves comparable performance to model-aware approaches with a decreased convergence rate.
Service robots benefit from encoding information in semantically meaningful ways to enable more robust task execution. Prior work has shown multi-relational embeddings can encode semantic knowledge graphs to promote generalizability and scalability, but only within a batched learning paradigm. We present Incremental Se…
Unsupervised scheme ranks sentences in text documents based on semantic importance.
problem Ranking sentences in text documents without labeled data.
method Extracts essential words and phrases, constructs semantic phrase and sentence graphs, applies PageRank, combines scores, and optimizes for topic diversity.
result SSR outperforms individual judges and compares favorably with combined rankings on benchmarks.
SPAT improves adversarial robustness by preserving semantics in adversarial training.
problem Adversarial examples often have different semantics than original data, introducing unintended biases.
method Semantics-preserving adversarial training (SPAT) that encourages pixel perturbation shared among all classes.
result SPAT improves adversarial robustness and achieves state-of-the-art results in CIFAR-10 and CIFAR-100.
Proposes CSG model to separate semantic and variation factors for OOD prediction.
problem Out-of-distribution examples cause conventional models to mix semantic and variation factors, leading to poor performance.
method Causal Semantic Generative model (CSG) based on causal reasoning, using variational Bayes for efficient learning and prediction.
result CSG can identify semantic factor and improve OOD prediction performance.
Building deep reinforcement learning agents that can generalize and adapt to unseen environments remains a fundamental challenge for AI. This paper describes progresses on this challenge in the context of man-made environments, which are visually diverse but contain intrinsic semantic regularities. We propose a hybrid …
New framework detects near vs. far out-of-distribution samples for AI safety.
problem Binary OOD detection fails to distinguish between semantically close and distant unknown risks.
method Ternary classification based on Low-Entropy Semantic Manifolds and Semantic Surprise Vector.
result Framework achieves state-of-the-art performance on ternary OOD detection task.
Embodied cognition states that semantics is encoded in the brain as firing patterns of neural circuits, which are learned according to the statistical structure of human multimodal experience. However, each human brain is idiosyncratically biased, according to its subjective experience history, making this biological s…
It is very useful to integrate human knowledge and experience into traditional neural networks for faster learning speed, fewer training samples and better interpretability. However, due to the obscured and indescribable black box model of neural networks, it is very difficult to design its architecture, interpret its …
Generative Adversarial Networks (GANs) have obtained extraordinary success in the generation of realistic images, a domain where a lower pixel-level accuracy is acceptable. We study the problem, not yet tackled in the literature, of generating semantic images starting from a prior distribution. Intuitively this problem…