Papers with CIA
AGrail: A Lifelong Agent Guardrail with Effective and Adaptive Safety Detection (2025.acl-long)
Copied to clipboard
| Challenge: | Existing defense agencies fail to adaptively and effectively mitigate these risks. |
| Approach: | They propose a lifelong agent guardrail that enhances LLM agent safety by enabling adaptive safety check generation, effective safety check optimization, and tool compatibility & flexibility. |
| Outcome: | The proposed agent guardrail achieves strong performance against task-specific and systemic risks and is transferable across different LLM agents’ tasks. |
CIA: Inferring the Communication Topology from LLM-based Multi-Agent Systems (2026.acl-long)
Copied to clipboard
| Challenge: | LLM-based multi-agent systems (MAS) have demonstrated remarkable capabilities in solving complex tasks. |
| Approach: | They propose a communication inference attack that constructs new adversarial queries to induce intermediate agents’ reasoning outputs and models their semantic correlations through the global bias disentanglement and LLM-guided weak supervision. |
| Outcome: | The proposed attack achieves an average AUC of 0.87 and a peak AUC up to 0.99, revealing the privacy risk in MAS. |
Cross-Lingual Auto Evaluation for Assessing Multilingual LLMs (2025.acl-long)
Copied to clipboard
Sumanth Doddapaneni, Mohammed Safi Ur Rahman Khan, Dilip Venkatesh, Raj Dabre, Anoop Kunchukuttan, Mitesh M Khapra
| Challenge: | Evaluating machine-generated text remains a challenge in NLP for non-English languages . current evaluation frameworks focus on English, revealing a gap in multilingual evaluations . |
| Approach: | They propose a cross-lingual auto evaluation framework that includes evaluator LLMs and a test set specifically designed for multilingual evaluation. |
| Outcome: | The proposed model aligns more closely with human judgments than proprietary models on non-English language evaluations. |
Retracing the Past: LLMs Emit Training Data When They Get Lost (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for extracting training data from large language models exhibit limited success . existing methods offer limited insight into the fundamental drivers of memorization leakage . |
| Approach: | They propose a framework for extracting memorized data by maximizing model uncertainty . they propose mismatched fine-tuning to weaken alignment and induce confusion . |
| Outcome: | The proposed attacks outperform baselines on unaligned and aligned LLMs . the proposed attacks exploit the model uncertainty of the input snippets induced by the model entropy spike . |