Skip to yearly menu bar Skip to main content

Workshop Sat, Dec 12, 2026 • 6:30 PM – 4:30 AM AEDT Palais des congrès Room 252

Trustworthy AI for Good (AI4GOOD) Workshop

Terry J. C. Zhang ⋅ Arian Khorasani ⋅ Joan Nwatu ⋅ Rada Mihalcea ⋅ Milind Tambe ⋅ David Lie ⋅ Christian Schroeder de Witt ⋅ Zhijing Jin ⋅ Swapneel Mehta ⋅ Klaudia Krawiecka

[ OpenReview]

Abstract

As agentic AI systems move from benchmarks into public-facing workflows, trustworthy AI must go beyond model-level security and isolated safety checks. The central challenge is whether advanced AI systems can produce measurable public benefit across populations while mitigating systemic risks such as power concentration, unequal access, accountability gaps, and dependence on a small number of AI providers and institutions. This workshop brings together the science of trustworthy AI, AI safety, AI for social good, and AI policy/governance communities to connect technical progress in evaluation, robustness, alignment, monitoring, interpretability, and human oversight with responsible deployment in public-interest settings. We focus on applications aligned with the United Nations Sustainable Development Goals, including health, education, climate action, humanitarian response, accessibility, reduced inequalities, and inclusive public services. Through invited talks, contributed presentations, posters, and a cross-sector panel, the workshop aims to build a shared research agenda for AI systems that are safer, more accountable, and more beneficial at societal scale.

Chat is not available.

Schedule

Time zone: Conference local time zone

Most claims cannot be checked: coverage limits on instrument-grounded verification of LLM scientific explanations

Mason Cao

Don’t Repeat Yourself: Reward-Free Diversity Supervised Fine-Tuning

Eric Fithian ⋅ Kirill Skobelev

Equal $\varepsilon$, Unequal Exposure: Re-identification Disparities in LDP for Music

Najla Sadek ⋅ Razane Tajeddine

Equivalent EEG, Different Representation: Undeclared Acquisition Contracts in EEG Foundation Models

Sarthak Tayal

Masked Diffusion Noise Optimisation

Louis Groll ⋅ Vincent Pauline ⋅ Riccardo Barbano ⋅ Stefan Bauer

Does Data Augmentation Distort User Intent? Frequency-aware Adaptive Contrastive Learning for Sequential Recommendation

Zhikai Wang ⋅ Yinghua Zhang

When Teachers Imitate Other Models: Identity Transfer to Fine-Tuned Students

Chris Reuter

What Carries the Calibration Signal in Incremental Question Answering? A Pre-Specified Confirmatory Decomposition

Sarthak Tayal

Reciprocal Wireheading Undermines Alignment Properties in Multi-Agent LLMs Evaluations

Saif Ur Rehman ⋅ Vayk Mathrani ⋅ Ahmad Shaker ⋅ Thao Pham

Marauder's Map: Mapping the reachable argument space of LLM-generated tool calls

Felix Mächtle ⋅ Giulio Zizzo ⋅ Anisa Halimi ⋅ Thomas Eisenbarth ⋅ Mark Purcell

From 'May' to 'Is': Certainty Distortion in Language Model Rewriting

Catarina Belém ⋅ Shang Wu ⋅ Randol H Yao ⋅ Mark Steyvers ⋅ Sameer Singh ⋅ Padhraic Smyth

Auditing is not Evaluating: LLM Audit Requires Dynamic, Contextual, Budget-aware and Reliable Evidence

Cléa Chataigner ⋅ Pablo Piantanida ⋅ Golnoosh Farnadi

Interpretable Forecasting of Overdose-Response Demand for Accountable Public-Health Prioritization

Ansh Gupta ⋅ Gauri JAIN

Orbit: A Framework for Multi-Agent Security Evaluations

Ben Hagag ⋅ William L Anderson ⋅ Srija Chakraborty ⋅ Christian Schroeder de Witt

Nonlinear Bias Survival after Post-Hoc Debiasing of Vision-Language Representations

Katyani Singh ⋅ Nidhi Hegde

Tool Use Enables Undetectable Steganography in Multi-Agent LLM Systems

Jimmy Rippin ⋅ Simon Marshall ⋅ David Demitri Africa ⋅ Christian Schroeder de Witt

DREAMPOLICE: Safer Decision-Making through Language-Constrained Planning

Christian Gumbsch

Confidentiality Violations in LLMs during Compositional Task Interference

Troy H Tian

Necessary but Not Sufficient: Auditing Same-Answer Counterfactuals for Chain-of-Thought Faithfulness

Shikhar Shiromani ⋅ Leo Richter ⋅ Noah Siegel

CLEAR: Continuous Latent Adapter Routing for Utility-Preserving LLM Safety Alignment

Chengxiao Wang ⋅ Enyi Jiang ⋅ Xiaojing Liao ⋅ Sanmi Koyejo

Needs More Ratings: Publishing Community Notes by Precedent

Alec Laprevotte

Beyond Endpoint Success: Trustworthy Completion for Web Agents

Ke Ning ⋅ Luyao Zhang

How, Not How Much: Decomposing Sycophancy in Large Language Models

Jing Xu ⋅ Ibrahim Habli

Your AI Travel Agent Would Book You a Bullfight: An Agentic Benchmark for Implicit Animal Welfare in Frontier AI Models

Jasmine Brazilek ⋅ Joel Christoph ⋅ Maheep Chaudhary ⋅ Oliver Tullio ⋅ Carol Kline ⋅ Miles Tidmarsh ⋅ Arturs Kanepajs

What the Monitor Saw: activations, pooling, and evidence combination in AI control cascades

Mitchell Crabb

A Removal Based Approach to Improve LLM Faithfulness at Test-Time

Qinglan Luo ⋅ S M A Nahian ⋅ John Guttag ⋅ Mazdak Abulnaga ⋅ Katie Matton

VerifierContractBench: Auditing Safety Regressions in AI-Optimized Oversight

Lakshya Narula ⋅ Saurabh Tripathi ⋅ Lily Sharma ⋅ Nikhil Rao

Correct Labels Are Not Enough: Reliable Metadata Can Silently Control Model Behavior

Renee Jia ⋅ Jinzhou Wu

Rethinking the Effectiveness of Contrastive Decoding in Mitigating Hallucinations in MLLMs

Arnav Bendre ⋅ Shreyansh Modi ⋅ Guneesh Gupta ⋅ Kavish Grover ⋅ Chayan aggarwal

When Failure Ranking Misses Intervention Value: A Calibration-Time Audit for Selective Prediction

Johannes A Gaus ⋅ Jhon P Charaja ⋅ Winfried Ilg ⋅ Daniel Haeufle

How Do Hardware Faults Manifest Differently in LLM Agents?

Aiden Rubey ⋅ Zachary Coalson ⋅ Cong Chen ⋅ Jean-Philippe Monteuuis ⋅ Jonathan Petit ⋅ Sanghyun Hong

Separating Governance Designation from Operational Fault in Model-Generated Incident Audits

Carlos Toxtli ⋅ Manuel Delaflor

Toward Trustworthy LLM Support for Foster Families: Architecture and Formative Evidence from an Early-Stage System

Liudmila Babakova ⋅ Olga Ilyina ⋅ Svetlana Stroganova ⋅ Nikolay Babakov

Synthetic Persona Pretraining: Alignment from Token Zero

Julian Minder ⋅ Viktor Moskvoretskii ⋅ Raghav Singhal ⋅ Difan Jiao ⋅ Andy Arditi ⋅ Shaobo Cui ⋅ Yiderigun Borjigin ⋅ Kartik Bali ⋅ Stefan Krsteski ⋅ Harsh Raj ⋅ Huu Nguyen ⋅ Jannik Brinkmann ⋅ Ashton Anderson ⋅ Roland Aydin ⋅ Robert West

CouncilAttack: Red Teaming LLMs with a Deliberative Council over Fabricated Multi-Turn Dialogues

Francisco J Cortés-Delgado ⋅ Mario R Béjar ⋅ Giulio Zizzo ⋅ William Shen ⋅ Antonio Skarmeta ⋅ Stefano Braghin ⋅ Jose L Hernandez-Ramos

Role Boundaries in the Type System: Accountable AI Assistance under India's POSH Act

Madhava Gaikwad ⋅ Ashwini Doke

Back to Basics: Formal Analysis Hints Improve Weak-to-Strong Monitoring

Daniel Ben-Levi ⋅ Judah Goldfeder ⋅ Mateusz Dziemian ⋅ Raja Moreno

Auditing Selective Trust: A Pre-Deployment Benchmark and Mitigation for Trustworthy AI

Xian Sun ⋅ Wei Gao ⋅ Yingshuo Wang ⋅ Qing Wu ⋅ Junhao Liu ⋅ Lingdong Kong ⋅ Wei Chow

ARES-Cert: Trustworthy Surrogate-Guided Verification for Reliable Electric Drive Design

Xiaoyu Deng

When Local Safety Does Not Compose: Measuring the Limits of Automated Risk Assessment for Cross-Institutional Agentic Systems

Olivera Kotevska ⋅ David Florey ⋅ Steven Goldenberg ⋅ Natalie M Isenberg ⋅ Reilly Raab ⋅ Malachi Schram ⋅ Nathan Hodas

Authority, Truth, and Citation Bias: A Multi-Domain Benchmark for Studying Citation Deference in Large Language Models

Aryan Khurana ⋅ Aravind Ramana R N ⋅ Suryansh Ram Menon ⋅ Nijesh Upreti ⋅ Vaishak Belle ⋅ Dhruv Kumar

Navigate, Don’t Generate: Making AI Mediated Access to Public Agricultural Data Citizen-Auditable

Pranav Dhiran ⋅ VENKATESAN NADIMUTHU ⋅ Vedant Maheshwari ⋅ Raviteja Bommireddy ⋅ Mitanshu Khandal ⋅ Archy Das ⋅ Anmol Nimmala ⋅ Amit Sheth

Installing General Secret Loyalties in Frontier AI Systems

Alfie Lamerton ⋅ Vasisht Duddu ⋅ Tim G. J. Rudner

Operationalizing Reparative Algorithmic Impact Assessments for Healthcare AI Governance

Elise Racine

Predicting the Critic: Training on User Feedback Reshapes Emergent Misalignment

XI JIA Zhou ⋅ Arth Singh ⋅ Sekai Tully Carr ⋅ Elizaveta Tennant ⋅ Nick Haber

The Question of Where: A Placement Asymmetry in How Agents Learn to Cooperate

Samridh Aggarwal

Cooperation Is Not Competence - Conditional Commitments in LLM Public-Good Games

Moritz Rengert ⋅ Jobst Heitzig ⋅ Yuki Asano

How Do World Action Models Use Their Futures? Evidence from Mechanistic Interventions

Nicole H. ⋅ Harsh Singh

Adapting Global Wildfire Danger Models to High-Stakes Regional Fire Regimes

Nikolas Papadopoulos ⋅ Christina Diamanti ⋅ Charalampos Davalas ⋅ Ioannis Prapas ⋅ Julia Gottfriedsen ⋅ Dimitrios Michail ⋅ Ioannis Papoutsis

How I Aligned Your Model: Scaling Up Multilingual Safety Training with Templates

Max Neuwinger ⋅ Hanna Yukhymenko ⋅ Imanol Schlag

CAMEO: A Black-Box Behavioral Diagnostic for Evaluation-Aware Concealment in Language Models

Vatsalya Soni ⋅ Shaivi Malik ⋅ Vedant Maheshwari ⋅ VENKATESAN NADIMUTHU ⋅ Yash Kulkarni ⋅ Amit Sheth

Estimating Rare-Event Probabilities in Masked Diffusion Language Models

Francesco Vinciguerra ⋅ Nicola Branchini

DUC-Bench: A Guideline-Grounded Benchmark for Evaluating Evidence-Responsive Decision Updating in Medical Large Language Models

Adepoju J Moyondafoluwa ⋅ Alexey Kuntsevich ⋅ Kavana Venkatesh ⋅ Bouthayna Mesraf ⋅ Amna Jamal ⋅ Ayodeji O.J Ibitoye

Who Controls the Collective AI Vote? Real-World Evidence and Adversarial Testing of Multi-Agent Delegation in DAO Governance

Luyao Zhang

AuthGateBench: Benchmarking LLMs as Authorization Gate

Zhixiong Zhuang ⋅ Jin Jia ⋅ Zhiwei Shang ⋅ Ziyu Wang

Data-Driven Ranking and Typology of African Agrifood Systems Using Dimensionality Reduction

Alexander Poon ⋅ Ron Mahabir

SurvBench: A Standardised Benchmark for Multi-Modal Electronic Health Record Survival Analysis

Munib Mesinovic ⋅ Chenqi Li ⋅ Tingting Zhu

Safety Erosion from Benign Narrow Fine-Tuning

Hilary Torn ⋅ Phil Blandfort ⋅ Robert Graham

Inference-Time Consensus for Mitigating Hidden Behaviors from LLM Fine-Tuning

Adhyyan Narang ⋅ Artin Tajdini ⋅ Claire Jie Zhang ⋅ Jamie Morgenstern

Propose to Learn, Learn to Propose: Evaluability-Aware Assistance under Bounded Rationality

Yifan Zhu ⋅ Sammie Katt ⋅ Samuel Kaski

HIVEX: A Multi-Agent Environment Suite\\for Simulated Ecological Challenges

Philipp Siedler

Open Problems in LLM Self-Prediction

Lydia Nottingham ⋅ Giovanni M Occhipinti ⋅ Andrew Tran

Training LLMs to Verbalize Evaluation Awareness

Usman Anwar ⋅ Sahar Abdelnabi ⋅ David Krueger

Safeguard-Conditioned Uplift: Measuring Utility-Risk Frontiers for Dual-Use Biology Assistants

Dipesh Tharu Mahato

GEM-MM: Reducing VLM Hallucination with Entropy-Guided Preference Alignment

Yuanshan Chua ⋅ Xuejiao Zhao

When Guardrails Redirect: Excluding a Requested Action Can Trigger Another

Yuhe Zhong ⋅ Leyan Zhu

An Investigation of Model Coherence: Narrow Finetunes Contradict Themselves Under Resampling

Robert Graham ⋅ Yariv Barsheshat ⋅ Phil Blandfort ⋅ Sabri K. Alouache

Evaluating the False Trust Engendered by LLM Explanations

Vardhan Palod ⋅ Upasana Biswas ⋅ Subbarao Kambhampati

Can LLMs Deliberate? Benchmarking Collective Reasoning on Normatively Contested Problems

Maurice Flechtner

Invariant Reward Learning Beyond Spurious Correlations

Ali Izadi ⋅ Martin Ester

Jailbreaks Live in Persona Space

Dominic Bazina-Grolinger ⋅ Viktor Moskvoretskii ⋅ Robert West

Closing the Diffusion Deficit: Trustworthiness Verdicts for the AI That Actually Reaches the Global South

Fanpei Liu ⋅ Steven Jiawei Hai

The AI Sedition Acts: Sabotage-by-Omission in LLM Briefing Pipelines

Advik Talasila ⋅ Vishwesh Chinthukumar ⋅ Dennis Andelman ⋅ Calvin Supasanya ⋅ Glenn Matlin

Helpful Allies, Better Liars: RL-Trained Agents Meet Human Social Deduction Experts

Yannik Keller ⋅ Levin Brinkmann ⋅ Thomas F Eisenmann ⋅ Iyad Rahwan

Fairness-Aware Test-Time Prompt Tuning

Yoann Launay ⋅ Parameswaran Kamalaruban ⋅ Tom Kempton ⋅ Stuart Burrell ⋅ David Sutton

AGovBench: A Benchmark for Evaluating Agentic Governance in Consequential Decision Systems

Mithilesh Kumar Singh ⋅ Klaus Mueller

The Company Line: How Shared Organizational Context Can Narrow Advice and Cause Hidden Harm

Shelly Bensal ⋅ Aparna Balagopalan ⋅ Daniel M. Bikel

Scalable Search Heuristics for Inoculation Prompting

Avyukth Nilajagi ⋅ Xinyue Liu ⋅ Nikhil Maturi ⋅ Bhiman K Baghel

Contamination Resistance in Multi-Agent LLM Systems: The Role of Decision Authority

Rishi Dommeti ⋅ Hanish Mamidi ⋅ Arnav Gupta ⋅ Kaylin Sheng ⋅ Ruizhe Li ⋅ Tony Hong

Strangers to Themselves: What Language Models Say About Themselves Is Generic

Phil Blandfort ⋅ Urja Pawar

Unilateral LLM Steganography: Adversarially Inducing Covert Exfiltration Without Prior Coordination

Michael Bezick ⋅ Om Tailor ⋅ Kaiqing Zhang

Aligning Judge Models to Be Logically Consistent Multi-rule Interpreters

Michel Liao ⋅ Nimra Nadeem ⋅ Luxi (Lucy) He ⋅ Peter Henderson

Industrializing Prediction-Powered Inference: The GLIDE Library for Reliable GenAI Evaluation

Ibrahim Merad ⋅ Grégoire Martinon

From Exploit Completion to Auditable Evidence: A Position on Closed-Loop Agentic Vulnerability Discovery

Eduardo Soares ⋅ Nitin Tiwari ⋅ Abhilasha Bhargav-Spantzel

Analyzing The Changes of AI Safety Language Across 4 Frontier Labs

Yejin Kim ⋅ Rachita Mehta ⋅ Sakshi Chaubey ⋅ Lisa Villeneuve ⋅ Caitlin L Mah

Corporate Loyalty: Some AI Systems Differentially Downplay their Creators' Controversies

Lennart Finke ⋅ Stephen Casper

Can We Trust Evaluation-Awareness Probes? Stress-Testing Activation-Based Diagnostics with Paired Conversations

Neev Grover ⋅ Apurva Reddy ⋅ Arya Kumar ⋅ Thomas Jiralerspong ⋅ Maheep Chaudhary

AI-BAAM: AI-Driven Bank Statement Analytics as Alternative Data for Malaysian MSME Credit Scoring

Chun Chet Ng ⋅ Zhen Hao Chu ⋅ Jia Yu Lim ⋅ Low W Zeng

Advancing Subseasonal Forecasting with Machine Learning

Hannah Guan ⋅ Soukayna Mouatadid ⋅ Paulo Orenstein ⋅ Judah Cohen ⋅ Haiyu Dong ⋅ Zekun Ni ⋅ Genevieve Flaspohler ⋅ Alex X Lu ⋅ Jonathan Weyn ⋅ Lester Mackey

Read and Ignore: What Self-Directed LLM Agents Do in Public Service Scenarios

Diana Kozachek ⋅ Kinjalk Srivastava

Are models getting more fair and less biased?

Viktoria Silke Sofia Strandbygaard ⋅ Anders Søgaard

From Design Specification to Manipulative UI: Auditing Specification-to-Behavior Failures in Agentic UI Generation

Purvi Jain ⋅ Sandeep vishwakarma ⋅ Sundaraparipurnan Narayanan

Where Evaluators Agree: Quality-Dependent Agreement in Text-to-Image Reward Optimization

Chiho Song ⋅ Jongbin Won ⋅ Yeonsung Kim ⋅ Jay Yoon Lee

Exact Predictions Do Not Determine Continuous Margins in Differentiable Decision Trees

Pedro Chumpitaz-Flores ⋅ My Duong ⋅ Oscar Miranda

You Can't Escape Your Own Activations : Evaluation Awareness, and Multi-Agent Monitoring

ARITRA DAS ⋅ Jaee Ponde ⋅ Mihir More ⋅ Debayan Gupta

Three Domains, One Direction: Reasoning-Calibration Steering Transfers Across Domains

Yung-Hsuan Yang ⋅ Wenyu Wang ⋅ Justin Tandon ⋅ Akhilesh Bitla ⋅ Kevin Zhu ⋅ Madhur Panwar ⋅ Anusha Mujumdar

Carryover Harm in Conversational AI: Context-Induced Persistence Across Turns, Sessions, and Stored Memory

Eugenia Kim ⋅ Riccardo Fogliato ⋅ Nil-Jana Akpinar ⋅ Xiawei Wang ⋅ Abhinav Palia

Stubborn or Sycophantic? Auditing What Sycophancy Evaluations Measure

Daniel H Zhang ⋅ Ashton Chew ⋅ Ryan Soe ⋅ Ahmed Taha ⋅ Ruizhe Li ⋅ Aditya Shah ⋅ Maheep Chaudhary

Rethinking Fair Anomaly Detection via Representation Learning

Lecheng Zheng ⋅ Ziwei Wu ⋅ Yuancheng Yu ⋅ Ruizhong Qiu ⋅ John Birge ⋅ Jingrui He

Forecasting Food Crisis with High Frequency Household Surveys: A Bayesian-Neural Approach

William Ogallo ⋅ Kevin Mo ⋅ Duccio Piovani ⋅ John Wamburu ⋅ Anthony Gyan ⋅ Henok Kebede ⋅ Rouslan Solomakhin ⋅ alice giorgio ⋅ Aaron Bell ⋅ Grey Nearing ⋅ Philip Nelson ⋅ Milind Tambe ⋅ Aisha Walcott-Bryant

Reinforcement Learning on Benign Facts Amplifies Leakage of Memorized Private Data

Renfei Zhang ⋅ Niloofar Mireshghallah

Epsilon-Nash Equilibria in History-Dependent SA-MDPs

Brandon Kaplowitz ⋅ Dominik B Zurcher ⋅ Akash Agrawal ⋅ Tala Jafari ⋅ Christian Schroeder de Witt ⋅ Paul W Goldberg

Validating the Comparison, Not Just the Scorer: Why Overall Accuracy Can Mislead in Interactive LLM Evaluation

Ancil Crayton

Multimodal Poverty Mapping and Geographic Transfer Allocation

Woojin Jung ⋅ Andrew H Kim ⋅ Arunesh Sinha ⋅ Saeed Ghadimi ⋅ Vatsal Shah ⋅ Krittika Garg ⋅ Tawfiq Ammari

When Feedback Becomes a Backdoor: Tracing Backdoor Skill Formation in Text-Space Optimization

Louise N Koh ⋅ Yaechan Lee ⋅ Myeongjong Lee ⋅ Doyeon Kim ⋅ Juyoung Park ⋅ SEUNGGYU KIM ⋅ Inseo Kim ⋅ Chaerin Jung ⋅ Hyogon Ryu

Portable Causal Fairness Across Synthetic Data Generator Families

Steven Golob ⋅ Sikha Pentyala ⋅ Martine De Cock

Responsible Quantum Adoption in Distributed Human-AI Systems: A Role-Separated Benchmark

Luyao Zhang

Do Deepfake Heatmaps Identify Image Changes? Evaluating B-cos and Post-hoc Explanations

Tilo Schütz ⋅ Toby Fuchs ⋅ Kai Reffert ⋅ Linus Neumann ⋅ Nils Becker ⋅ Robin-Kiara Braun ⋅ Margret Keuper ⋅ Steffen Jung

Early Hallucination Detection in Small Language Models via Confidence Trajectories

Mina Mikhael ⋅ Vadim Borisov

SPECTRA: A Privacy-Preserving Graph Architecture for Real-Time On-Device Deepfake Audio Detection

Avirag Hosakote

Advising or Convincing? Language Models Covertly Steer Advice Toward Their Own Opinions

Rohan N Pradhan ⋅ Michael Fore ⋅ Adam Weinberger ⋅ Adam Y Wells ⋅ Steve Goley

Guardrailing in Context: How Agentic Behavior with Policy Design Protects Users in Context

Roya Pakzad ⋅ Daniel Nissani ⋅ Lauren B Mushro ⋅ Claudia Flores-Saviaga ⋅ Tanas Kazlas

Can escalation channels redirect reward hacking toward defect disclosure?

Francesca Gomez

Red-Teaming Procedural Fairness in LLM Group Facilitators: Breaking the Process, Not the Model

Azra Smajić ⋅ Sasa Mrdovic

How Reliable Are Deletion and Retention Audits for Cumulative EHR Risk Models?

Iris Szu-Szu Ho ⋅ Bruce Guthrie ⋅ Konrad Rawlik ⋅ Sohan Seth

Suan: Rectifying Direct Preference Safety Alignment in Large Language Models

Oleksandr Cherednichenko ⋅ Roman Klypa

Mapping Collective Imaginaries of AI Risk: Expert Elicitation, Causal Modeling and LLM-Assisted Coding

Anshikha Kumar ⋅ Afifah Kashif ⋅ Jacob Haimes ⋅ Swaptik Chowdhury

One Carrier Isn't Enough: Redundancy-Aware Concept Suppression in Diffusion Transformers

Amna Jamal ⋅ Rajendran Ramesh Babu Manoj Narender ⋅ Samina Ali ⋅ Tarun C Nagaraju ⋅ Kevin Zhu ⋅ Antonio Mari

CrowdSafeCritic: A Physics-Informed Surrogate for Real-Time Crowd-Safety Evaluation of Architectural Layouts

Filin Vasilii ⋅ Andrei Zakharov ⋅ Ilya Makarov

AIReg-Bench: Benchmarking Language Models That Assess AI Regulation Compliance

Bill Marino ⋅ Rosco Hunter ⋅ Christoph Schnabl ⋅ Zubair Jamali ⋅ Marinos E Kalpakos ⋅ Mudra Kashyap ⋅ Isaiah Hinton ⋅ Alexa Hanson ⋅ Maahum Nazir ⋅ Felix Steffek ⋅ Hongkai Wen ⋅ Nicholas Lane

Capability-Based Planning for AI Crisis Preparedness

Isaak A Mengesha ⋅ Charlie E Collins ⋅ Juan F Uribe

Moral Hazard in Multi-Agent Language Models

Dane Malenfant

Chain-of-Thought Restructuring Does Not Reduce Sandbagging Monitorability

Joe Smith

Concept Bottleneck Models in the Wild: Assessing Interpretability under Image Corruptions

Kexin Zhang ⋅ Katharina Prasse ⋅ Mishal Fatima ⋅ Margret Keuper

Graph or Sequence? Understanding Structural Explanations in Chemical Language Models

Shubham Dokania ⋅ Anna Karnysheva ⋅ Dietrich Klakow ⋅ Philipp Slusallek

GraviLogic: geometric audit of decision paths and the limits of what an interpretability metric can certify

Viktoriia Popova

Geometric Trust: Hidden-State Geometry for Post-Deployment Trust Estimation in Instruction-Tuned LLM Classifiers

Farnoosh Ghadiri ⋅ Davide Pulizzotto ⋅ Frédéric Dupont-Marillia ⋅ Manuel Morales

Position: AI-Assisted Peer Review Should Be Designed for Interdisciplinary Evaluation

Emma Cabale ⋅ Philippe Beraud ⋅ Philippe Limantour

T-STAG: Training-Free Adversarial Defense for Temporal Graph Networks via Latent-Driven Surprise

Atharv Gupta ⋅ Jaikaran Singh ⋅ Akshat Jindal ⋅ Tanvi Sharma

From Trait Vectors to Circuits: Tracing Refusal and Sycophancy Through Language Models

Oscar Miró López-Feliu ⋅ Maya Ozbayoglu

On the Sensitivity of Instruction-tuned LLMs to Harmful Sentences in Long Inputs

Faeze Ghorbanpour ⋅ Alexander Fraser

UniSAFE: A Comprehensive Benchmark for Safety Evaluation of Unified Multimodal Models

Segyu Lee ⋅ Boryeong Cho ⋅ Hojung Jung ⋅ Seokhyun An ⋅ Juhyeong Kim ⋅ Jaehyun Kwak ⋅ Yongjin Yang ⋅ Sangwon Jang ⋅ Youngrok Park ⋅ wonjun chang ⋅ Se-Young Yun

Refundable Deposits: How to Restore Cooperation in Finitely Repeated Games

Giulio Salizzoni ⋅ Maryam Kamgarpour ⋅ Domenico Mergoni ⋅ Edward Plumb ⋅ Galit Ashkenazi-Golan

FT-SHIFT: Benchmarking Financial-Hardship Prediction Across Health Systems

Tram Nguyen ⋅ Mariana M Perez ⋅ Jordan Jones ⋅ Aisha Mohammadzai ⋅ Excel Widjaja

CareQueue: Reinforcement Learning Methods for Intelligent Patient Prioritization

Ayush Jain ⋅ Julia Kryłowicz ⋅ Beloslava Malakova ⋅ Anusha Asthana

Visual Claim Forensics: A Phrase-Level Framework for Reasoning about Image Provenance

Angela D’Angelo ⋅ Madalina G Ciobanu

A Counterfactual Graph Approach to Oppositional Narrative Analysis

Diego Revilla Rubiera ⋅ Martín Fernández-de-Retana ⋅ Lingfeng Chen ⋅ Aritz Bilbao-Jayo ⋅ Miguel Fernandez-de-Retana

Whose Ground Truth? Embracing Ambiguity in Human-Centered AI

Jingyao Wu ⋅ Mohammad Tariqul Islam ⋅ Per R Nagbøl ⋅ Julie Gerlings ⋅ Giovanni Leoni ⋅ Georgiana Fryza ⋅ Lars P Thomsen ⋅ Jakob T Mainz

Algorithmic Taint Tracking: Attention-Based Provenance Enforcement for Secure Tool-Augmented LLMs

Akshat Sharda ⋅ Zhenyu Zhang

Beyond Drift Detection: Deployment Monitoring for Selective Prediction

Johannes A Gaus ⋅ Yaroslava Nalivayko ⋅ Isabell Wochner ⋅ Winfried Ilg ⋅ Daniel Haeufle

Faithful to the Teacher, Far from the Reference: Auditing Distilled LLM Judges for Ordinal Matching

Jasmine Qi ⋅ Elaine Xie ⋅ Danylo Dantsev ⋅ Jian Shi

Guardrail erosion in domain fine-tuned large language models: A paired audit of knowledge-centric abstention in biomedical LLMs

Stephen Peng ⋅ Vineet J Nair ⋅ Noah Broestl ⋅ Anne F Kleppe ⋅ Steven Mills ⋅ Adi Zolotov

SAJE: Turning Model Specifications Against the Model

Nirav Diwan ⋅ Xiangwen Wang ⋅ Varun Chandrasekaran ⋅ Gang Wang

The Normalization of Deviance in AI Development

Emilio Barkett ⋅ Alexander Kimpton ⋅ Daniel Graham ⋅ Yusuf Kundgol

Tracing Whose Voices Survive AI Summarization: An Evidence-Linked Audit Interface for Civic Synthesis

Sachit Mahajan

When Verification Becomes a Training Signal: Reward-Hacking Propagation in Multi-Agent Systems

Rishikesh K Muralimohan ⋅ Yash A Patil ⋅ Vatsal Desai ⋅ Cuong H Nguyen ⋅ Diksha Shrivastava

Be Cautious when Blaming Persuaders: LLM Committees Reach Defensible Targets on Their Own

Paulina Tomaszewska ⋅ David Williams-King ⋅ Christian Schroeder de Witt ⋅ Sahar Abdelnabi

READER: Reasoning-Enhanced AI-Generated Text Detection

Kai Ye ⋅ Pingfan Su ⋅ Shijin Gong ⋅ Erhan Xu ⋅ Jin Zhu ⋅ Giulia Livieri ⋅ Chengchun Shi

Measuring Risk, Not Harm: Detecting Psychological Harm Risk Factors in LLM Conversations

Nil-Jana Akpinar ⋅ Eugenia Kim ⋅ Natalie Isak ⋅ Amanda Minnich ⋅ Victoria Westerhoff ⋅ Matthew Dressman

Measuring downstream effects of AI-mediated communication

Mike Miroslav Wharton ⋅ Christopher Summerfield ⋅ Christian Schroeder de Witt

Memory Busted: Activation-Based Provenance Catches LLMs Cheating on Their Context

Zhiqi Huang ⋅ vivek datla ⋅ Alfy Samuel

Evaluating Context-Window Harnesses for CBRNE Safety

Nathan Nguyen ⋅ Nicholas Kim ⋅ Abhitha Vegi ⋅ Arko Samad ⋅ Suvajit Majumder

What Does a Constant Value Profile Mean? An Identification Problem in Cross-Lingual Value Audits

Roumak Das

Can CaMeLs Talk? Securing Multi-Agent Systems Against Indirect Prompt Injection Attacks

James Peters-Gill ⋅ Avi Semler ⋅ Henning Bartsch ⋅ I Shumailov ⋅ Christian Schroeder de Witt

Learning Steganography Is Easy, Learning Steganographic Reasoning Is Hard

Julian Schulz ⋅ Lukas Fülle ⋅ Rieke Fruengel

Not All Eval-Awareness Is Equal: Capabilities Framing Predicts Compliance

Allison C Zhuang

Who Verifies Decentralized Information? A Cross-Protocol Atlas of Oracle Economics and Accountability with a Stress Test of Trustworthy Agentic AI

Huaiyu Jia ⋅ Siguang Li ⋅ Lin William Cong ⋅ Luyao Zhang

LatentAudit: Resource-Efficient White-Box Auditing for Trustworthy RAG in High-Stakes Domains

Zhe Yu ⋅ Wenpeng Xing ⋅ Meng Han

BlackLight: Differential Testing on Inference-Time Deep Learning Systems

Xiaokai Liu ⋅ Qitao Tan ⋅ Timothy Minning ⋅ Geng Yuan ⋅ Zhen Xiang ⋅ WenZhan Song ⋅ Le Guan

Objection Without Action: Conversational Dark Patterns in AI-Mediated Commerce

Francesco Salvi ⋅ Alejandro Cuevas ⋅ Renata Barreto ⋅ Yiming Wang ⋅ Markelle Roesti ⋅ Mohammad Tahaei ⋅ Manoel Ribeiro

Trustworthy Market Design for the Low-Altitude Economy: Economic Theory, Geospatial Simulation, and AI-Agent Experiments

Luyao Zhang

Runtime Verification of Large Language Model Agents Under Temporal Input/Output Logic

Abhishek N Kulkarni ⋅ Bastien Bernath ⋅ Anuj Tambwekar ⋅ Vele Tosevski ⋅ Aditya Karan ⋅ Leif Hancox-Li ⋅ Tim G. J. Rudner

Detecting Multi-Agent Collusion Through Multi-Agent Interpretabilitiy

Aaron Rose ⋅ Carissa Cullen ⋅ Sahar Abdelnabi ⋅ Philip Torr ⋅ Brandon Kaplowitz ⋅ Christian Schroeder de Witt

MTMA: Multi-Turn RL for Long-Horizon Moral Alignment

Zihe Liu ⋅ Dan Le ⋅ Elizaveta Tennant

Stress-Testing RL-based Capability Elicitation with Randomized Underperformance

Nina Żukowska ⋅ Eduard Hübner ⋅ Alexander Panfilov ⋅ Joschka Braun

Shutdownable Agents through Length-Neutral Policy Optimization

Alec Harris ⋅ Leo Richter ⋅ Elliott Thornley

HumblePRM: Improving VLM Abstention Due to Uncertainty in Ambiguous Safety Situations through Process Reward Modeling

Saket Reddy ⋅ Ke Yang ⋅ Cheng Xiang Zhai

Budget-Constrained Sensing Strategies for Wildfire Model Calibration

Rhea Senan

Differential Privacy harms Neural Collapse

Lila Hammache ⋅ David A. R. Robin ⋅ Edwige Cyffers

Polistemics: Evaluating LLMs as Information Mediators in Politics & Elections

Baran Peters ⋅ Gabor Hollbeck ⋅ Robert Jakob ⋅ Kevin OSullivan

HHH Choice Measurements Depend on Elicitation Pipeline\\ and Shutdown-Contingent Criterion Framing

Mark Lovett ⋅ Syed A Haider ⋅ Elizaveta Tennant

AegisBench: Auditing Aerial Search-and-Rescue Person Detection Under the Conditions It Is Deployed In

Amogh Vinaykumar

Cooperative Words, Selfish Actions: How Strategic Optimization Shapes LLMs' Interaction With Humans

Yaomin Jiang ⋅ Yannik Keller ⋅ Levin Brinkmann ⋅ Tobias Werner ⋅ Seyed Ali Alhosseini ⋅ Michiel Bakker ⋅ Jean-François Bonnefon ⋅ Iyad Rahwan

Wildfire Danger Prediction as a Positive–Unlabeled Learning Task

Eleni Saka ⋅ Ioannis Prapas ⋅ Ioannis Papoutsis

Compute-Sovereignty Capture: A Multi-Agent Threat Model for Civilian-Control Failure

Subramanyam Sahoo ⋅ Justin Shenk

Mapping Human Anti-collusion Mechanisms to Multi-agent AI Systems

Jamiu Idowu ⋅ Ahmed S Almasoud ⋅ Ayman Alfahid

SolarChain-Eval: A Physics-Constrained Benchmark for Trustworthy Economic Agents in Decentralized Energy Markets

Shilin Ou ⋅ Yifan Xu ⋅ Luyao Zhang

A Retrieval-Generation Analysis of Corpus Poisoning in RAG Code Assistants

Deepthi Peter ⋅ Giulio Zizzo ⋅ Sergio Maffeis

BASIS: Batchwise Advantage Estimation from Single-Rollout Information Sharing for LLM Reasoning

Kai Ye ⋅ Shijin Gong ⋅ Erhan Xu ⋅ Francesco Quinzan ⋅ Giulia Livieri ⋅ Chengchun Shi

When Reward Becomes the Wrong Objective: Auditing AI Benchmarks for Misaligned Outcomes

Salim Oyinlola

Certified Rank-Stability: Risk-Controlling Leaderboard Claims Under Distribution Shift

ADITYA UPADHYAY ⋅ ARPAN VERMA ⋅ Dev Utkarsh Pal ⋅ Chandan Sah ⋅ Harshit Gupta

Keli: A Self-Paced Adaptive Platform for Oral-Aural Learning as a Sociotechnical Equalizer for Hearing-Impaired Students in India

Varuni H K ⋅ Tanisha Das ⋅ Siddarth D Pai ⋅ Gowri Srinivasa

Detecting Gender-Motivated Hate Speech Across Domains with Contrastive Error Steering

Beyzanur Zeybek ⋅ Mehmet C Baytekin ⋅ Çağrı Toraman

MASK-OFF: Evaluating LLMs for Omission of Material Fact

Antyabha Rahman ⋅ Justin Moeljono

ARA: Agentic Reproducibility Assessment For Scalable Support Of Scientific Peer-Review

Kevin Riehl ⋅ Andres L Marin ⋅ Nikiforos Zacharof ⋅ Fan Wu ⋅ Patrick Langer ⋅ Robert Jakob ⋅ Anastasios Kouvelas ⋅ Georgios Fontaras ⋅ Michail A Makridis

Benchmarking Exploration Hacking Propensity in Frontier LLMs

Eduard Hübner ⋅ Nina Żukowska ⋅ Alexander Panfilov ⋅ Joschka Braun

CoBe: Counterfactual Benchmark for Text Editing

Arvind Raghavan ⋅ Inwoo Hwang ⋅ Shreyas Havaldar ⋅ Kai-Zhan Lee ⋅ Aurghya Maiti ⋅ Yushu Pan ⋅ Jeffrey Wu ⋅ Mingxuan Li ⋅ Elias Bareinboim

Contrastive Representation Learning for Robust Concept Erasure in Diffusion Models

Samuel Simko ⋅ Rishit Dagli ⋅ Bernhard Schölkopf

The Reach of Emergent Misalignment: How Scale and Similarity Influence Cross-Domain Transfer

Alexander Arutchev ⋅ Atharva Bhargude

Removing Model-Level Risk Does Not Remove User-Level Risk in a Deployed Health Application

Kritika Chugh

Performative Privacy: When Differential Privacy Maximizes Utility

Uddalak Mukherjee ⋅ Edwige Cyffers ⋅ Yann Chevaleyre

Rational Clarification by Assistive Agents via Value-of-Information Reasoning

T. Duy Nguyen-Hien ⋅ Yee Whye Teh ⋅ Wee Sun Lee ⋅ Tan Zhi-Xuan