Cover Image

Safe But Not Sorry: Reducing Over-Conservatism in Safety Critics via Uncertainty-Aware Modulation

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Bethell, Daniel, Gerasimou, Simos, Calinescu, Radu, Imrie, Calum
Format:	Preprint
Published:	2025
Subjects:	Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Safe Reinforcement Learning in Black-Box Environments via Adaptive Shielding
by: Bethell, Daniel, et al.
Published: (2024)

Learning to Navigate Under Imperfect Perception: Conformalised Segmentation for Safe Reinforcement Learning
by: Bethell, Daniel, et al.
Published: (2025)

Robust Uncertainty Quantification Using Conformalised Monte Carlo Prediction
by: Bethell, Daniel, et al.
Published: (2023)

Robust Adversarial Quantification via Conflict-Aware Evidential Deep Learning
by: Barker, Charmaine, et al.
Published: (2025)

Guided Uncertainty Learning Using a Post-Hoc Evidential Meta-Model
by: Barker, Charmaine, et al.
Published: (2025)

Conformal Safety Shielding for Imperfect-Perception Agents
by: Scarbro, William, et al.
Published: (2025)

Uncertainty Quantification for Deep Regression using Contextualised Normalizing Flows
by: Marco, Adriel Sosa, et al.
Published: (2025)

DeepKnowledge: Generalisation-Driven Deep Learning Testing
by: Missaoui, Sondess, et al.
Published: (2024)

Mind the Prompt: Self-adaptive Generation of Task Plan Explanations via LLMs
by: Vázquez, Gricel, et al.
Published: (2026)

Formal Synthesis of Uncertainty Reduction Controllers
by: Carwehl, Marc, et al.
Published: (2024)

Verification and External Parameter Inference for Stochastic World Models
by: Calinescu, Radu, et al.
Published: (2025)

Assuring the Safety of Reinforcement Learning Components: AMLAS-RL
by: Imrie, Calum Corrie, et al.
Published: (2025)

Sparsity-based Safety Conservatism for Constrained Offline Reinforcement Learning
by: Cho, Minjae, et al.
Published: (2024)

A Safety Modulator Actor-Critic Method in Model-Free Safe Reinforcement Learning and Application in UAV Hovering
by: Qi, Qihan, et al.
Published: (2024)

Handling Long-Term Safety and Uncertainty in Safe Reinforcement Learning
by: Günster, Jonas, et al.
Published: (2024)

Uncertainty-Aware Tabular Prediction: Evaluating VBLL-Enhanced TabPFN in Safety-Critical Medical Data
by: Ramalingam, Madhushan
Published: (2025)

Dyna-Style Safety Augmented Reinforcement Learning: Staying Safe in the Face of Uncertainty
by: Eisele, Artur, et al.
Published: (2026)

Safe Domain Randomization via Uncertainty-Aware Out-of-Distribution Detection and Policy Adaptation
by: Danesh, Mohamad H., et al.
Published: (2025)

Formally Guaranteed Control Adaptation for ODD-Resilient Autonomous Systems
by: Vázquez, Gricel, et al.
Published: (2026)

Learning Fairer Representations with FairVIC
by: Barker, Charmaine, et al.
Published: (2024)

Accelerating Policy Synthesis in Large-Scale MDPs via Hierarchical Adaptive Refinement
by: Evangelidis, Alexandros, et al.
Published: (2025)

Optimal Parameter Adaptation for Safety-Critical Control via Safe Barrier Bayesian Optimization
by: Wang, Shengbo, et al.
Published: (2025)

Uncertainty-Aware Predictive Safety Filters for Probabilistic Neural Network Dynamics
by: Frauenknecht, Bernd, et al.
Published: (2026)

Better Safe than Sorry: Pre-training CLIP against Targeted Data Poisoning and Backdoor Attacks
by: Yang, Wenhan, et al.
Published: (2023)

Assessing the Chemical Intelligence of Large Language Models
by: Runcie, Nicholas T., et al.
Published: (2025)

Molecular Representations for Large Language Models
by: Runcie, Nicholas T., et al.
Published: (2026)

Uniformly Safe RL with Objective Suppression for Multi-Constraint Safety-Critical Applications
by: Zhou, Zihan, et al.
Published: (2024)

Safe Urban Traffic Control via Uncertainty-Aware Conformal Prediction and World-Model Reinforcement Learning
by: Chandra, Joydeep, et al.
Published: (2026)

Safety Controller Synthesis for Collaborative Robots
by: Gleirscher, Mario, et al.
Published: (2020)

On Safety in Safe Bayesian Optimization
by: Fiedler, Christian, et al.
Published: (2024)

Safe LoRA: the Silver Lining of Reducing Safety Risks when Fine-tuning Large Language Models
by: Hsu, Chia-Yi, et al.
Published: (2024)

Long-Horizon Model-Based Offline Reinforcement Learning Without Explicit Conservatism
by: Ni, Tianwei, et al.
Published: (2025)

CAPSULE: Control-Theoretic Action Perturbations for Safe Uncertainty-Aware Reinforcement Learning
by: Narava, Rahul, et al.
Published: (2026)

Dissecting Sample Hardness: A Fine-Grained Analysis of Hardness Characterization Methods for Data-Centric AI
by: Seedat, Nabeel, et al.
Published: (2024)

SorryDB: Can AI Provers Complete Real-World Lean Theorems?
by: Letson, Austin, et al.
Published: (2026)

SafeAug: Safety-Critical Driving Data Augmentation from Naturalistic Datasets
by: Mo, Zhaobin, et al.
Published: (2025)

Safe Langevin Soft Actor Critic
by: Keswani, Mahesh, et al.
Published: (2026)

Emergent Risk Awareness in Rational Agents under Resource Constraints
by: Ornia, Daniel Jarne, et al.
Published: (2025)

Safe Continual Reinforcement Learning under Nonstationarity via Adaptive Safety Constraints
by: Tomashevskiy, Timofey
Published: (2026)

Compositional Conservatism: A Transductive Approach in Offline Reinforcement Learning
by: Song, Yeda, et al.
Published: (2024)