Imagen de Portada

How Many Iterations to Jailbreak? Dynamic Budget Allocation for Multi-Turn LLM Evaluation

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Feldman, Shai, Romano, Yaniv
Formato:	Preprint
Publicado:	2026
Materias:	Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Robust Conformal Prediction Using Privileged Information
por: Feldman, Shai, et al.
Publicado: (2024)

Conformal Prediction with Corrupted Labels: Uncertain Imputation and Robust Re-weighting
por: Feldman, Shai, et al.
Publicado: (2025)

Calibrated Predictive Lower Bounds on Time-to-Unsafe-Sampling in LLMs
por: Davidov, Hen, et al.
Publicado: (2025)

Valid Best-Model Identification for LLM Evaluation via Low-Rank Factorization
por: Tolochinsky, Elad, et al.
Publicado: (2026)

Automating Deception: Scalable Multi-Turn LLM Jailbreaks
por: Kumarappan, Adarsh, et al.
Publicado: (2025)

Label Noise Robustness of Conformal Prediction
por: Einbinder, Bat-Sheva, et al.
Publicado: (2022)

Semi-Supervised Hypothesis Testing by Betting on Predictions
por: Tenzer, Yaniv, et al.
Publicado: (2026)

Multi-Turn Jailbreaks Are Simpler Than They Seem
por: Yang, Xiaoxue, et al.
Publicado: (2025)

Adaptive Budget Allocation in LLM-Augmented Surveys
por: Ye, Zikun, et al.
Publicado: (2026)

LLM Defenses Are Not Robust to Multi-Turn Human Jailbreaks Yet
por: Li, Nathaniel, et al.
Publicado: (2024)

Not All Turns Matter: Credit Assignment for Multi-Turn Jailbreaking
por: He, Zhida, et al.
Publicado: (2026)

Attention-Aware GNN-based Input Defense against Multi-Turn LLM Jailbreak
por: Huang, Zixuan, et al.
Publicado: (2025)

Multi-Task Combinatorial Bandits for Budget Allocation
por: Ge, Lin, et al.
Publicado: (2024)

ZEBRA: Zero-shot Budgeted Resource Allocation for LLM Orchestration
por: Hamri, May, et al.
Publicado: (2026)

Online Learning with Improving Agents: Multiclass, Budgeted Agents and Bandit Learners
por: Ashkezari, Sajad, et al.
Publicado: (2026)

Star Elastic: Many-in-One Reasoning LLMs with Efficient Budget Control
por: Taghibakhshi, Ali, et al.
Publicado: (2026)

Knowledge-Driven Multi-Turn Jailbreaking on Large Language Models
por: Li, Songze, et al.
Publicado: (2026)

Not All Turns Are Equally Hard: Adaptive Thinking Budgets For Efficient Multi-Turn Reasoning
por: Jali, Neharika, et al.
Publicado: (2026)

Uncertainty Quantification and Data Efficiency in AI: An Information-Theoretic Perspective
por: Simeone, Osvaldo, et al.
Publicado: (2025)

Protected Test-Time Adaptation via Online Entropy Matching: A Betting Approach
por: Bar, Yarin, et al.
Publicado: (2024)

Efficient Budget Allocation for Large-Scale LLM-Enabled Virtual Screening
por: Li, Zaile, et al.
Publicado: (2024)

Building Math Agents with Multi-Turn Iterative Preference Learning
por: Xiong, Wei, et al.
Publicado: (2024)

Robust Conformal Outlier Detection under Contaminated Reference Data
por: Bashari, Meshi, et al.
Publicado: (2025)

Mitigating Many-Shot Jailbreaking
por: Ackerman, Christopher M., et al.
Publicado: (2025)

Semi-Supervised Risk Control via Prediction-Powered Inference
por: Einbinder, Bat-Sheva, et al.
Publicado: (2024)

Jailbreak Attack Initializations as Extractors of Compliance Directions
por: Levi, Amit, et al.
Publicado: (2025)

Pivotal Auto-Encoder via Self-Normalizing ReLU
por: Goldenstein, Nelson, et al.
Publicado: (2024)

Learning a Continue-Thinking Token for Enhanced Test-Time Scaling
por: Ringel, Liran, et al.
Publicado: (2025)

SINR-Aware Deep Reinforcement Learning for Distributed Dynamic Channel Allocation in Cognitive Interference Networks
por: Cohen, Yaniv, et al.
Publicado: (2024)

LAVa: Layer-wise KV Cache Eviction with Dynamic Budget Allocation
por: Shen, Yiqun, et al.
Publicado: (2025)

AutoAdv: Automated Adversarial Prompting for Multi-Turn Jailbreaking of Large Language Models
por: Reddy, Aashray, et al.
Publicado: (2025)

Aligning LLMs Toward Multi-Turn Conversational Outcomes Using Iterative PPO
por: Jiang, Daniel R., et al.
Publicado: (2025)

Testing For Distribution Shifts with Conditional Conformal Test Martingales
por: Shaer, Shalev, et al.
Publicado: (2026)

Jailbreaking Large Language Models in Infinitely Many Ways
por: Goldstein, Oliver, et al.
Publicado: (2025)

Cascaded Transfer: Learning Many Tasks under Budget Constraints
por: Campagne, Eloi, et al.
Publicado: (2026)

X-Teaming: Multi-Turn Jailbreaks and Defenses with Adaptive Multi-Agents
por: Rahman, Salman, et al.
Publicado: (2025)

Reinforcing Multi-Turn Reasoning in LLM Agents via Turn-Level Reward Design
por: Wei, Quan, et al.
Publicado: (2025)

TROJail: Trajectory-Level Optimization for Multi-Turn Large Language Model Jailbreaks with Process Rewards
por: Xiong, Xiqiao, et al.
Publicado: (2025)

Hidden Representation Clustering with Multi-Task Representation Learning towards Robust Online Budget Allocation
por: Wang, Xiaohan, et al.
Publicado: (2025)

Synthetic-Powered Multiple Testing with FDR Control
por: Lee, Yonghoon, et al.
Publicado: (2026)