Overconfidence

Follow this topic Ask about this topic

Momentum

5 papers in the last four weeks, up 67% on the four weeks before. 0.0% of all new papers.

Jul 13Week of Sep 28

Latest papers 47

All topics

Guard Models Are Overconfident Where Base Models Are Uncertain
Sep 29, 2026Jonghyun Hong, MinJae Jung, Minwoo KimLarge Language Model Safety Confidence Calibration
DATUMO INC.
When Confidence Rises Too Early: Detecting Shortcut Reasoning via Premature Answer Commitment
Sep 28, 2026Zhaohan Zhang, Junjie Liu, Chengzhengxu Li +6Reasoning Skills Overconfidence
Queen Mary University of London · Tongyi Lab, Alibaba Group · Xi’an Jiaotong University
When World Models Lie: Adaptive Safety Analysis Under Wrong Imaginations
Sep 28, 2026John Cao, Somil BansalSafety Filters Latent World Models
Department of Aeronautics and Astronautics, Stanford University
Calibration as a First-Class Criterion in LLM Evaluation
Sep 22, 2026Mario Sanz-Guerrero, Katharina von der WenseLarge Language Model Evaluation Overconfidence
Johannes Gutenberg University Mainz, Germany · University of Colorado Boulder, USA
Rethinking Verbalized Confidence for LLM-as-a-Judge: A Compatibility Shift on Post-2025 Proprietary Models
Sep 11, 2026Yu-Chung HsiaoLlm-As-A-Judge Overconfidence
Cisco Systems
Are You Sure You're Sure? On the Impact of Instruction Tuning on Confidence and Lexical Diversity
Aug 13, 2026Irina Proskurina, Mayank Kumar, Oyindolapo O. KomolafeInstruction Tuning Instruction-Tuned Models
Cohere Labs Community · Laboratoire Hubert Curien, UMR CNRS 5516, Saint-Étienne, France · School of Computer Science Engineering and Technology (SCSET), Bennett University, Greater Noida, India +1
Revisiting Overestimation Bias Problem of Q-learning: Settling Large Discrete Action Space via Action Intersection
Aug 13, 2026Pu Li, Tao Tan, Hong Xie +2Q-Learning Action Space
Chongqing Institute of Green and Intelligent Technology, Chongqing, China · University of Science and Technology of China, Hefei, Anhui, China
When Confidence Fails: Overconfidence in LLMs under Uncertainty and Missing Clinical Information
Aug 10, 2026Maryam Tahermazandarani, Adnan Mahmood, Fahmida Islam +1Overconfidence Uncertainty
School of Computing, Macquarie University, Sydney, NSW 2109, Australia
Beyond Post-Hoc Temperature Scaling: Bilevel Optimization for LLM Calibration
Aug 7, 2026Ruochen Jin, Zhanliang Wang, Zongyu Dai +2Overconfidence Loss Function
Dartmouth College, Hanover, NH · University of Pennsylvania, Philadelphia, PA · National University of Singapore, Singapore
GeoReward: Mitigating Contextual Variable Overestimation in Vision-Language Models for Cross-Market Preference Prediction
Aug 5, 2026Shuo Liu, Huixiang Cai, Weiru Zhang +1Ads Geogs-Slam
Alibaba International Digital Commerce Group
Revisiting TD Target Aggregation under Uncertainty in Q-Learning
Aug 4, 2026Lipeng Zu, Xiaonan ZhangDeep Q-Networks Q-Learning
Computer Science Department, Florida State University, Tallahassee, Florida, USA
Belief-Guided Decision Making with Uncertainty Gating in the Game of Go
Jul 29, 2026Mehrad Yaghoubi, Azam Bastanfard, Abbas Jalilvand +1Monte Carlo Tree Search Chess
Department of Computer Engineering, Ka.C, Islamic Azad University, Karaj, Iran
Collaborative Weighting with Pessimistic Critic for Mitigating Overestimation in Off-Policy Reinforcement Learning
Jul 29, 2026Gong Gao, Xiao Lai, Ziqi Xie +3Soft Actor-Critic Pessimism
School of Computer Science and Technology, Tongji University, 200092, Shanghai, China · Shanghai University of Engineering Science, 201620, Shanghai, China
Priors learned from legacy reconstructions inherit undetectable overconfidence
Jul 23, 2026Ali Siahkoohi, Sina AlemohammadBayesian Inverse Problems Confidence Intervals
Department of Computer Science University of Central Florida Orlando, FL, USA · Department of Electrical and Computer Engineering The University of Texas at Austin Austin, TX, USA
Judge-dependent safety gains and model-specific helpfulness costs of evidence-sufficiency prompting in clinical LLMs
Jul 20, 2026Koyar AfrasyabLarge Language Model Judges Overconfidence
Kinvectum AB, Sweden
The Severance Problem: LLMs are Unaware of the Person Beyond the Prompt
Jul 15, 2026Dor Litvak, Liu LeqiLarge Language Models Fail Artificial Intelligence Assistants
The University of Texas at Austin
When Prompts Ignore Structure: Graph-Based Attribute Reasoning for Calibrated VLMs
Jul 8, 2026Tanay Sodha, Aditya Sharma, Ramya Hebbalaguppe +2Vision-Language Model Adaptation Recent Vision-Language Models
Department of Computer Science and Information Systems, Birla Institute of Technology and Science, Pilani, India · TCS Research, New Delhi, India
ConflictScore: Identifying and Measuring How Language Models Handle Conflicting Evidence
Jun 24, 2026Siyi Liu, Aaron Halfaker, Dan Roth +1Evidence Conflict Truth
University of Pennsylvania · Microsoft
One-Step Generalization Ratio Guided Optimization for Domain Generalization
Jun 15, 2026Sumin Cho, Dongwon Kim, Kwangsu KimValue Alignment Overconfidence
Department of Computer Science and Engineering, University of Sungkyunkwan, Suwon, Korea.
Confidence Laundering in Agent Systems: Why Uncertainty Needs a Latent Carrier
Jun 9, 2026Kaiwen Shi, Zheyuan Zhang, Han Bao +2Uncertain Uncertainty
University of Notre Dame
GUI-AC: Enhancing Continual Learning in GUI Agents
Jun 9, 2026Can Lin, Tao Feng, Hangjie Yuan +3Graphical User Interface Agents Reinforcement Fine-Tuning
Beijing University of Posts and Telecommunications · Tsinghua University · Zhejiang University +1
GIScholarBench: Benchmarking LLM Overconfidence in GIS Research
Jun 6, 2026Zongrng Li, Mingzheng Yang, Lei Zou +8Scholar Information Database Overconfidence
Department of Geography Texas A&M University College Station, TX, USA · Google Mountain View, CA, USA · Department of Landscape Architecture and Urban Planning Texas A&M University College Station, TX, USA
Calibrating Overconfidence Without Sacrificing Confidence: Probe-Conditioned Head Intervention for LLMs
Jun 2, 2026Ke Li, Chongzhe Zhang, Zifan Zeng +3Overconfidence Instruction-Tuned Models
Huawei Heisenberg Research Center · EPFL · TU Berlin +1
Large Language Models Are Overconfident in Their Own Responses
Jun 2, 2026Mario Sanz-Guerrero, Manuel Mager, Katharina von der WenseInstruction-Tuned Models Overconfidence
∇Johannes Gutenberg University Mainz, Germany · ♠University of Colorado Boulder, USA
Seeing Isn't Knowing: Do VLMs Know When Not to Answer Spatial Questions (and Why)?
May 28, 2026Yue Zhang, Zun Wang, Han Lin +3Spatial Reasoning Visual Evidence
UNC Chapel Hill · Google Research
Conservative neural posterior estimation via distributionally robust training
May 27, 2026William Laplante, Yuga Hikida, Charita Dellaporta +2Distributionally-Robust Optimization Overconfidence
Department of Statistical Science, University College London, UK · The Alan Turing Institute, UK · Department of Computer Science, Aalto University, Finland
Reward Bias Substitution: Single-Axis Bias Mitigations Redirect Optimization Pressure
May 27, 2026Max Lamparth, Daniel Fein, Andreas Haupt +2Progress Reward Modeling Mitigation
Stanford University · University of Pennsylvania
Are We Overconfident in Models and Results for Semi-Supervised 3D Medical Image Segmentation?
May 25, 2026Jun Li, Ziwei Qin3D Medical Image Segmentation Semantic Segmentation
Institute of Systems Science and Technology, School of Electrical Engineering, Southwest Jiaotong University, China.
Understanding and Mitigating Premature Confidence for Better LLM Reasoning
May 23, 2026Jingchu Gai, Guanning Zeng, Christina Baek +4LLM Reasoning Strategies Overconfidence
1Carnegie Mellon University · 2Tsinghua University
Regret Analysis of Retry-Based Bandits
May 20, 2026Bingkui Tong, Junpei Komiyama, Soichiro Nishimori +1Stochastic Multi-Armed Bandits Regret
Mohamed bin Zayed University of Artificial Intelligence · RIKEN AIP · The University of Tokyo
Generative-Evaluative Agreement: A Necessary Validity Criterion for LLM-Enabled Adaptive Assessment
May 19, 2026Grandee Lee, Yue Wang, Che Yee Lye +1Computerized Adaptive Testing Rubrics
Singapore University of Social Sciences
Sign-Separated Asymmetric Finite-Time Error Analysis of Q-Learning
May 15, 2026Donghwan LeeQ-Learning Value Functions
Department of Electrical Engineering Korea Advanced Institute of Science and Technology (KAIST) Daejeon 34141, South Korea
The Silent Vote: Improving Zero-Shot LLM Reliability by Aggregating Semantic Neighborhoods
May 10, 2026Sanket Badhe, Priyanka Tiwari, Deep ShahGumbel-Softmax Relaxation Overconfidence
Google
Beyond ECE: Calibrated Size Ratio, Risk Assessment, and Confidence-Weighted Metrics
May 3, 2026Fernando Martin-Maroto, Nabil Abderrahaman, Gonzalo G. de PolaviejaOverconfidence Confidence Calibration
Mathematics of Behavior and Intelligence Lab. Champalimaud Research. Avenida de Brasília, Lisbon · Algebraic AI Calle Castelló, 117, Madrid
Calibration Drift Under Reasoning: How Chain-of-Thought Budgets Induce Overconfidence in Large Language Models
Apr 24, 2026Prakul Sunil Hiremath, Harshit R. HiremathLLM Reasoning Strategies Overconfidence
Department of Computer Science and Engineering Visvesvaraya Technological University, Belagavi · Department of Computer Science and Business System SG Balekundri Institute of Technology, Belagavi
From If-Statements to ML Pipelines: Revisiting Bias in Code-Generation
Apr 23, 2026Minh Duc Bui, Xenia Heilmann, Mattia Cerrato +2Large Language Model Bias Code Generation
Johannes Gutenberg University Mainz, Germany · Universidad Iberoamericana, Ciudad de Mexico · University of Colorado Boulder, USA
You Don't Need Public Tests to Generate Correct Code
Apr 23, 2026Kaushitha Silva, Srinath PereraTest Generation Code Generation
WSO2 Santa Clara, CA, USA
Revisiting Catastrophic Forgetting in Continual Knowledge Graph Embedding
Apr 21, 2026Gerard Pons, Carlos Escolano, Besim Bilalli +1Knowledge Graph Embeddings Catastrophic Forgetting
Data-intensive Technologies and Knowledge Systems (DTAK), Department of Service and Information System Engineering, Universitat Politècnica de Catalunya, Barcelona, 08034, Catalonia, Spain · Intelligent Data Science and Artificial Intelligence Research Center (IDEAI), Department of Computer Science, Universitat Politècnica de Catalunya, Barcelona, 08034, Catalonia, Spain
QuantSightBench: Evaluating LLM Quantitative Forecasting with Prediction Intervals
Apr 17, 2026Jeremy Qin, Maksym AndriushchenkoLarge Language Model Forecasting Frontier Models
ELLIS Institute Tübingen · Max Planck Institute for Intelligent Systems · Tübingen AI Center
Overconfidence and Calibration in Medical VQA: Empirical Findings and Hallucination-Aware Mitigation
Apr 2, 2026Ji Young Byun, Young-Jin Park, Jean-Philippe Corbeil +1Medical Visual Question Answering Confidence Calibration
Johns Hopkins University, School of Medicine · Massachusetts Institute of Technology · Microsoft Healthcare & Life Sciences
DamageArbiter: A Multimodal Arbitration Framework for Disaster Damage Assessment from Street-View Imagery
Mar 16, 2026Yifan Yang, Lei Zou, Wenjing Gong +6Post-Disaster Damage Assessment Multimodal Model
Department of Geography, Texas A&M University, College Station, USA · Department of Landscape Architecture & Urban Planning, Texas A&M University, College Station, USA · Department of Industrial and Systems Engineering, University of Florida, Gainesville, USA +2
Diagnosing Corruption-Induced Reliability Failures in Vision-Language Models
Nov 24, 2025Xiangjie Sui, Songyang Li, Hanwei Zhu +3Overconfidence
City University of Macau, Macau SAR, China · Jiangxi University of Finance and Economics, Nanchang, China · Nanyang Technological University, Singapore
When Guessing is Rewarded: Rethinking Language Model Evaluation with Distributional Uncertainty Scoring
Oct 5, 2025Thomas F BurnsUncertainty Score Large Language Model Evaluation
Aleph Alpha Research
Ensemble Elastic DQN: A Step Dependent Ensemble Approach for Reducing Overestimation in Deep Value-Based Reinforcement Learning
Jun 6, 2025Adrian Ly, Richard Dazeley, Peter Vamplew +2Deep Q-Networks Value Functions
Towards Understanding The Calibration Benefits of Sharpness-Aware Minimization
May 29, 2025Chengli Tan, Yubo Zhou, Haishan Ye +7Sharpness-Aware Minimization Recalibration
Northwestern Polytechnical University · SGIT AI Lab, State Grid Corporation of China · Xi’an Jiaotong University +2
Popular but Wrong: Understanding and Mitigating LLM Overconfidence through Knowledge Popularity
May 23, 2025Shiyu Ni, Keping Bi, Jiafeng Guo +1Overconfidence Large Language Models Fail
CAS Key Lab of Network Data Science and Technology, ICT, CAS · State Key Laboratory of AI Safety · University of Chinese Academy of Sciences
Understanding and Mitigating Under-Confidence in GNNs from the Final Layer
May 16, 2025Jincheng Huang, Jie Xu, Xiaoshuang Shi +3Graph Neural Networks Confidence Estimation
School of Computer Science and Engineering, University of Electronic Science and Technology of China · Southeast University

Sep 29, 2026, cs.LGJ/K move · Enter open · S save

Guard Models Are Overconfident Where Base Models Are Uncertain

Jonghyun Hong, MinJae Jung, Minwoo Kim

DATUMO INC.

Large Language Model Safety Confidence Calibration Overconfidence Divergence

Open paper AskPDF arXiv

Guard models are used as safety classifiers, with confidence scores driving downstream moderation decisions. We evaluate five guard models for prompt classification and find that although several are nearly calibrated on clean inputs, adversarial attacks degrade their calibration by an order of magnitude, turning false negatives into high-confidence errors indistinguishable from correct detections. Comparing each guard with its corresponding base LM, we find that uncertainty signals often remain available, with the base model typically expressing uncertainty on the same inputs where the guard fails. Layer-wise analyses localize this guard-base divergence to later layers, where guard models exhibit sharper safe/unsafe separation and lower-rank representations, while adversarial harmful inputs lie closer to the clean-safe region. These findings highlight a mismatch between guard confidence and base model uncertainty under attack.

Momentum

Related topics

Latest papers 47