📰 Workshop on Trustworthy NLP 2026: all accepted papers

41 papers · the most cited first · ← Workshop on Trustworthy NLP 2026

1. Through a Compressed Lens: Investigating The Impact of Quantization on Factual Knowledge RecallQianli Wang, Mingyang Wang, Nils Feldhus et al.2 citations · our summary
2. A Systematic Comparison between Extractive Self-Explanations and Human Rationales in Text Classification ↗Stephanie Brandl, Oliver Eberle1 citation
3. KoLegalQA: A Korean Legal QA Dataset for Trustworthy and Explanation-Grounded Legal AI ↗Yongtae Lee, Surin Lee, Sumin Kim et al.1 citation
4. Purdah and Patriarchy: Evaluating and Mitigating South Asian Biases in Open-Ended Multilingual LLM GenerationsMamnuya Rinki, Chahat Raj, Anjishnu Mukherjee et al.1 citation · our summary
5. A Systematic Taxonomy of Failure Modes in Retrieval-Augmented Generation Systems ↗Anupama Garani
6. Authorization-First Retrieval: Enforcing Least Privilege in Multi-Agent RAG SystemsRohith Namboothiriour summary
7. ChatbotManip: a Dataset to Facilitate Evaluation and Oversight of Manipulative Chatbot Behaviour ↗Jack Luigi Henry Contro, Simrat Deol, Martim Brandao et al.
8. ClaimCLAIRE: A Trust-Aware Multi-Component Fact-Checking Agent for Open-World Claims ↗Xinman Liu, Mayank Sharma
9. Coercion Suppression Increases Preference Hallucinations via a Deceptive Bypass in $K$-Level Negotiation Agents ↗Jihye Kim
10. Context Misleads LLMs: The Role of Context Filtering in Maintaining Safe Alignment of LLMs ↗Jinhwa Kim, Ian Harris
11. Controllable Pareto Trade-off between Fairness and Accuracy ↗Yongkang Du, Jieyu Zhao, Yijun Yang et al.
12. Deactivating Refusal Triggers: Understanding and Mitigating Overrefusal in Safety Alignment ↗Zhiyu Xue, Zimo Qi, Guangliang Liu et al.
13. Did You Forget What I Asked? Prospective Memory Failures in Large Language ModelsAvni Mittalour summary
14. Domain-Dependent Safety Behavior in Open-Weight LLMs: An Empirical Study Across Seven Ethical Domains ↗Zacharie Bugaud
15. Don't Want Your LLM to Recommend Nuclear Strike? Try Asking It in JapaneseRian Touchentour summary
16. Evaluating Cross-Lingual Behavior and Consistency of Multimodal Large Language ModelsHao Wang, Pinzhi Huang, Daisuke Kawaharaour summary
17. GateKD: Confidence-Gated Closed-Loop Distillation for Robust ReasoningKasidit Sermsri, Teerapong Panboonyuenour summary
18. Geometric Deviation as an Unsupervised Pre-Generation Reliability Signal: Probing LLM Representations for Answerability ↗Yucheng Du
19. Ghost Context: Measuring Cross-Context Interference in Long-Context Language ModelsRohith Namboothiriour summary
20. Guiding Giants: Lightweight Controllers for Weighted Activation Steering in LLMs ↗Amr Hegazy, Mostafa Elhoushi, Amr Alanwar
21. Hair-Trigger Alignment: Black-Box Evaluation Cannot Guarantee Post-Update Alignment ↗Yavuz Faruk Bakman, Duygu Nur Yaldiz, Salman Avestimehr et al.
22. Improving the Faithfulness of LLM-based Abstractive Summarization with Span-level Unlikelihood Training ↗Sicong Huang, Qianqi Yan, Shengze Wang et al.
23. Lexical Familiarity Predicts Processing Depth for Nonliteral Language in Large Language Models ↗Lang-Ching Yeh, Yu-Chieh Wang, Shu-Kai Hsieh
24. Linear Probes Detect Task Format, Not Reasoning Mode in Language Model Hidden StatesSubramanyam Sahoo, Vinija Jain, Aman Chadha et al.our summary
25. Multilingual Steering by Design: Multilingual Sparse Autoencoders and Principled Layer Selection ↗Yusser Al Ghussin, Daniil Gurgurov, Tanja Baeumel et al.
26. PII Jailbreaking in LLMs via Activation Steering Reveals Personal Information LeakageKrishna Kanth Nakka, Xue Jiang, Dmitrii Usynin et al.our summary
27. Quantifying LLM Safety Degradation Under Repeated Attacks Using Survival AnalysisZvi Topolour summary
28. ReacTOD: Bounded Neuro-Symbolic Agentic NLU for Zero-Shot Dialogue State Tracking ↗Yanjun Lin, Zimo Xiao, Kartik Natarajan et al.
29. SURGELLM: Rethinking Multi-Task Evaluation through Task-Aware Feature Gating with Class-Balanced NormalizationNoor Islam S. Mohammad, Ulug Bayazitour summary
30. Single-Layer Activation Edits Easily Corrupt Factual Recall but Rarely Repair It ↗Zacharie Bugaud
31. Teaching People LLM's Errors and Getting it RightNathan Stringham, Fateme Hashemi Chaleshtori, Xinyuan Yan et al.our summary
32. The Conservative AI: Diagnosing Hold Bias and Reliability Limits in Persona-Based Monetary Policy Simulation ↗Giyong Kim, Sojung Kim
33. The Geometry of Refusal: Linear Instability in Safety-Aligned LLMsShivam Ratnakar, Kartikeya Vatsour summary
34. The Halo Effect and Language Takeover: Spatiotemporal Attention Decay Explains Vision-Language Model Failures in Simple Visual Counting ↗Haochen Zhao, Sujian Li
35. Toward Dialect-Aware Safety Evaluation for Arabic Large Language Models ↗Wajdi Zaghouani
36. Truth or Dare: Analyzing LLM Susceptibility to External Evidence of Varying FactualityHan-Yu Su, Kuan-Yu Chu, Yung-Hui Li et al.our summary
37. Uncertainty-Aware Proxy Attribute Reasoning for Reliable Media Bias Detection ↗Chin-Po Chen, Jeng-Lin Li, Ming-Ching Chang
38. Understanding the Effects of Safety Unalignment on Reasoning- and Instruction-Tuned Large Language Models ↗John Timothy Halloran
39. What are They Thinking? Delineation, Probing, and Tracking of Concepts in LLMs ↗Mohamed Abdelwahab, Michelle Yu Collins, Sihan Chen et al.
40. Why is “Chicago” Predictive of Deceptive Reviews? Using LLMs to Discover Language Phenomena from Lexical CuesJiaming Qu, Mengtian Guo, Yue Wangour summary
41. With a Grain of SALT: Are LLMs Fair Across Social Dimensions? ↗Samee Arif, Zohaib Khan, Maaidah Kaleem Butt et al.

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/venues/trustnlp/2026/papers/