Publications

2026

1. Last Translation Benchmark

   Vilém Zouhar, Niyati Bafna, Mukund Choudhary, Maike Züfle, and 1 more author

   arXiv preprint arXiv:2609.04173, 2026

   arXiv Bib

   @article{zouhar2026last,
     title = {Last Translation Benchmark},
     author = {Zouhar, Vil{\'e}m and Bafna, Niyati and Choudhary, Mukund and Z{\"u}fle, Maike and others},
     journal = {arXiv preprint arXiv:2609.04173},
     year = {2026},
   }

2. Improved Robustness Against Indirect Prompt Injection Can Be Built into LLMs

   M. Özdinçer, Samuel Simko, Bernhard Schölkopf, and Zhijing Jin

   In Conference on Language Modeling (COLM), 2026

   COLM 2026 Bib

   Accepted to COLM 2026

   @inproceedings{ozdincer2026injection,
     title = {Improved Robustness Against Indirect Prompt Injection Can Be Built into LLMs},
     author = {{\"O}zdin{\c c}er, M. and Simko, Samuel and Sch{\"o}lkopf, Bernhard and Jin, Zhijing},
     booktitle = {Conference on Language Modeling (COLM)},
     year = {2026},
   }

3. Training with Honeypots: Reshaping How LLMs Fail Under Adversarial Attacks

   Samuel Simko, Punya Syon Pandey, Zhijing Jin, and Bernhard Schölkopf

   In International Conference on Machine Learning (ICML), 2026

   ICML 2026 Bib HTML Code

   Accepted to ICML 2026

   @inproceedings{simko2026honeypots,
     title = {Training with Honeypots: Reshaping How LLMs Fail Under Adversarial Attacks},
     author = {Simko, Samuel and Pandey, Punya Syon and Jin, Zhijing and Sch{\"o}lkopf, Bernhard},
     booktitle = {International Conference on Machine Learning (ICML)},
     year = {2026},
   }

4. Keeping Agents on Track: Using Context-Induced Failures as a Learning Signal

   M. Özdinçer, Samuel Simko, Bernhard Schölkopf, and Zhijing Jin

   In AI4GOOD Workshop at ICML, 2026

   ICML 2026 Workshop Bib

   Accepted to the AI4GOOD Workshop at ICML 2026

   @inproceedings{ozdincer2026agents,
     title = {Keeping Agents on Track: Using Context-Induced Failures as a Learning Signal},
     author = {{\"O}zdin{\c c}er, M. and Simko, Samuel and Sch{\"o}lkopf, Bernhard and Jin, Zhijing},
     booktitle = {AI4GOOD Workshop at ICML},
     year = {2026},
   }

5. Position: Safe Models Do Not Guarantee Safe Societies: The Case for Sociopolitical Risk

   David Guzman Piedrahita, Changling Li, Dave Banerjee, Terry Jingchen Zhang, and 7 more authors

   In International Conference on Machine Learning (ICML), Position Paper Track, 2026

   ICML 2026 Spotlight Bib HTML

   Accepted as a spotlight to the ICML 2026 Position Paper Track

   @inproceedings{piedrahita2026position,
     title = {Position: Safe Models Do Not Guarantee Safe Societies: The Case for Sociopolitical Risk},
     author = {Guzman Piedrahita, David and Li, Changling and Banerjee, Dave and Zhang, Terry Jingchen and Blin, Kevin and Simko, Samuel and Pandey, Punya Syon and Strauss, Irene and Mihalcea, Rada and Sch{\"o}lkopf, Bernhard and Jin, Zhijing},
     booktitle = {International Conference on Machine Learning (ICML), Position Paper Track},
     year = {2026},
   }

6. TamperBench: Systematically Stress-Testing LLM Safety Under Fine-Tuning and Tampering

   Saad Hossain, Tom Tseng, Punya Syon Pandey, Samanvay Vajpayee, and 7 more authors

   arXiv preprint arXiv:2602.06911, 2026

   KDD 2026 Oral arXiv Bib

   Accepted as an oral to the KDD 2026 Datasets and Benchmarks Track

   @article{hossain2026tamperbench,
     title = {TamperBench: Systematically Stress-Testing LLM Safety Under Fine-Tuning and Tampering},
     author = {Hossain, Saad and Tseng, Tom and Pandey, Punya Syon and Vajpayee, Samanvay and Kowal, Matthew and Nonta, Nayeema and Simko, Samuel and Casper, Stephen and Jin, Zhijing and Pelrine, Kellin and Rambhatla, Sirisha},
     journal = {arXiv preprint arXiv:2602.06911},
     year = {2026},
   }

2025

1. 

   Improving Large Language Model Safety with Contrastive Representation Learning

   Samuel Simko, Mrinmaya Sachan, Bernhard Schölkopf, and Zhijing Jin

   arXiv preprint arXiv:2506.11938, 2025

   EMNLP 2025 arXiv Bib

   Accepted as a poster to EMNLP Main 2025

   @article{simko2025improving,
     title = {Improving Large Language Model Safety with Contrastive Representation Learning},
     author = {Simko, Samuel and Sachan, Mrinmaya and Sch{\"o}lkopf, Bernhard and Jin, Zhijing},
     journal = {arXiv preprint arXiv:2506.11938},
     year = {2025},
   }

2. Causal AI Scientist: Facilitating Causal Data Science with Large Language Models

   Vishal Verma, Sawal Acharya, Samuel Simko, Devansh Bhardwaj, and 5 more authors

   In AI for Science Workshop at NeurIPS, 2025

   NeurIPS 2025 Workshop Bib HTML

   Accepted to the NeurIPS 2025 AI for Science Workshop

   @inproceedings{verma2025causal,
     title = {Causal AI Scientist: Facilitating Causal Data Science with Large Language Models},
     author = {Verma, Vishal and Acharya, Sawal and Simko, Samuel and Bhardwaj, Devansh and Haghighat, Anahita and Sachan, Mrinmaya and Janzing, Dominik and Sch{\"o}lkopf, Bernhard and Jin, Zhijing},
     booktitle = {AI for Science Workshop at NeurIPS},
     year = {2025},
   }

3. 

   Accidental Vulnerability: Factors in Fine-Tuning that Shift Model Safeguards

   Punya Syon Pandey, Samuel Simko, Kellin Pelrine, and Zhijing Jin

   arXiv preprint arXiv:2505.16789, 2025

   IASEAI 2026 arXiv Bib

   Accepted to IASEAI’26

   @article{pandey2025accidental,
     title = {Accidental Vulnerability: Factors in Fine-Tuning that Shift Model Safeguards},
     author = {Pandey, Punya Syon and Simko, Samuel and Pelrine, Kellin and Jin, Zhijing},
     journal = {arXiv preprint arXiv:2505.16789},
     year = {2025},
   }