Maksym Andriushchenko

Identifiers

• name variant Maksym Andriushchenko 0.60 · backfill

Papers (40)

1. What Shapes Emergent Misalignment? Insights from Training Dynamics, Model Priors, and Data cs.AI · 2026 · author #4
2. Decomposing and Measuring Evaluation Awareness cs.LG · 2026 · author #6
3. FutureSim: Replaying World Events to Evaluate Adaptive Agents cs.LG · 2026 · author #7
4. Europe and the Geopolitics of AGI: The Need for a Preparedness Plan cs.CY · 2026 · author #11
5. Instrumental Choices: Measuring the Propensity of LLM Agents to Pursue Instrumental Behaviors cs.AI · 2026 · author #3
6. Characterizing the Consistency of the Emergent Misalignment Persona cs.AI · 2026 · author #3
7. QuantSightBench: Evaluating LLM Quantitative Forecasting with Prediction Intervals cs.LG · 2026 · author #2
8. Claudini: Autoresearch Discovers State-of-the-Art Adversarial Attack Algorithms for LLMs cs.LG · 2026 · author #6
9. Skill-Inject: Measuring Agent Vulnerability to Skill File Attacks cs.CR · 2026 · author #4
10. Helpful to a Fault: Measuring Illicit Assistance in Multi-Turn, Multilingual LLM Agents cs.CL · 2026 · author #4
11. Monitoring Decomposition Attacks in LLMs with Lightweight Sequential Monitors cs.CR · 2025 · author #4
12. Exploring Memorization and Copyright Violation in Frontier LLMs: A Study of the New York Times v. OpenAI 2023 Lawsuit cs.LG · 2024 · author #4
13. AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents cs.LG · 2024 · author #1
14. Does Refusal Training in LLMs Generalize to the Past Tense? cs.CL · 2024 · author #1
15. Improving Alignment and Robustness with Circuit Breakers cs.LG · 2024 · author #6
16. Is In-Context Learning Sufficient for Instruction Following in LLMs? cs.CL · 2024 · author #2
17. Competition Report: Finding Universal Jailbreak Backdoors in Aligned LLMs cs.CL · 2024 · author #5
18. Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attacks cs.CR · 2024 · author #1
19. JailbreakBench: An Open Robustness Benchmark for Jailbreaking Large Language Models cs.CR · 2024 · author #4
20. Long Is More for Alignment: A Simple but Tough-to-Beat Baseline for Instruction Fine-Tuning cs.CL · 2024 · author #2
21. Scaling Compute Is Not All You Need for Adversarial Robustness cs.LG · 2023 · author #3
22. Critical Influence of Overparameterization on Sharpness-aware Minimization cs.LG · 2023 · author #3
23. Why Do We Need Weight Decay in Modern Deep Learning? cs.LG · 2023 · author #2
24. Layer-wise Linear Mode Connectivity cs.LG · 2023 · author #2
25. Transferable Adversarial Robustness for Categorical Data via Universal Robust Embeddings cs.LG · 2023 · author #2
26. Sharpness-Aware Minimization Leads to Low-Rank Features cs.LG · 2023 · author #1
27. A Modern Look at the Relationship between Sharpness and Generalization cs.LG · 2023 · author #1
28. SGD with Large Step Sizes Learns Sparse Features cs.LG · 2022 · author #1
29. Towards Understanding Sharpness-Aware Minimization cs.LG · 2022 · author #1
30. ARIA: Adversarially Robust Image Attribution for Content Provenance cs.CV · 2022 · author #1
31. On the effectiveness of adversarial training against common corruptions cs.LG · 2021 · author #2
32. RobustBench: a standardized adversarial robustness benchmark cs.LG · 2020 · author #2
33. Understanding and Improving Fast Adversarial Training cs.LG · 2020 · author #1
34. Sparse-RS: a versatile framework for query-efficient sparse black-box adversarial attacks cs.LG · 2020 · author #2
35. On the Stability of Fine-tuning BERT: Misconceptions, Explanations, and Strong Baselines cs.LG · 2020 · author #2
36. Square Attack: a query-efficient black-box adversarial attack via random search cs.LG · 2019 · author #1
37. Provably Robust Boosted Decision Stumps and Trees against Adversarial Attacks cs.LG · 2019 · author #1
38. Why ReLU networks yield high-confidence predictions far away from the training data and how to mitigate the problem cs.LG · 2018 · author #2
39. Logit Pairing Methods Can Fool Gradient-Based Attacks cs.LG · 2018 · author #2
40. Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulation cs.LG · 2017 · author #2

Mentions

• 2506.10949 #4 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2311.17539 #3 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2410.09024 #1 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2407.11969 #1 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2405.19874 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2404.02151 #1 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2412.06370 #4 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2310.04415 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2404.01318 #4 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2406.04313 #6 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2404.14461 #5 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2402.04833 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2307.06966 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2312.13131 #3 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2306.04064 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2305.16292 #1 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2302.07011 #1 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2210.05337 #1 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2206.06232 #1 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2202.12860 #1 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2006.12834 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2103.02325 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2010.09670 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2006.04884 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2007.02617 #1 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 1912.00049 #1 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 1906.03526 #1 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 1812.05720 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 1810.12042 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 1705.08475 #2 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2606.20814 #4 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2603.24511 #6 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2605.23055 #6 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2602.16346 #4 · arxiv_oai · confidence 0.70 Maksym Andriushchenko
• 2602.20156 #4 · arxiv_oai · confidence 0.70 Maksym Andriushchenko

Frequent Coauthors

• Nicolas Flammarion 18 shared papers
• Francesco Croce 9 shared papers
• Matthias Hein 8 shared papers
• Edoardo Debenedetti 4 shared papers
• Vikash Sehwag 3 shared papers
• Aditya Varre 2 shared papers
• Andy Zou 2 shared papers
• Anietta Weckauff 2 shared papers
• Dan Hendrycks 2 shared papers
• Derek Duenas 2 shared papers
• Dietrich Klakow 2 shared papers
• Hao Zhao 2 shared papers
• Jonas Geiping 2 shared papers
• Justin Wang 2 shared papers
• Klim Kireev 2 shared papers
• Marius Mosbach 2 shared papers
• Matt Fredrikson 2 shared papers
• Maxwell Lin 2 shared papers
• Sahar Abdelnabi 2 shared papers
• Yuchen Zhang 2 shared papers