Image de couverture de livre

Token Statistics Reveal Conversational Drift in Multi-turn LLM Interaction

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Hafez, Wael, Nazeri, Amir
Format:	Preprint
Publié:	2026
Sujets:	Computation and Language Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Information-Theoretic Framework for Self-Adapting Model Predictive Controllers
par: Hafez, Wael, et autres
Publié: (2026)

Entropy-Based Non-Invasive Reliability Monitoring of Convolutional Neural Networks
par: Nazeri, Amirhossein, et autres
Publié: (2025)

The Informational Cost of Agency: A Bounded Measure of Interaction Efficiency for Deployed Reinforcement Learning
par: Hafez, Wael, et autres
Publié: (2026)

MAC: A Multi-Agent Framework for Interactive User Clarification in Multi-turn Conversations
par: Acikgoz, Emre Can, et autres
Publié: (2025)

Mutual Information Tracks Policy Coherence in Reinforcement Learning
par: Reid, Cameron, et autres
Publié: (2025)

Drift No More? Context Equilibria in Multi-Turn LLM Interactions
par: Dongre, Vardhan, et autres
Publié: (2025)

On the Multi-turn Instruction Following for Conversational Web Agents
par: Deng, Yang, et autres
Publié: (2024)

A Mathematical Theory of Agency and Intelligence
par: Hafez, Wael, et autres
Publié: (2026)

Stop Listening to Me! How Multi-turn Conversations Can Degrade LLM Reliability
par: Guo, Kevin H., et autres
Publié: (2026)

Asymmetric Actor-Critic for Multi-turn LLM Agents
par: Jiang, Shuli, et autres
Publié: (2026)

Defensive M2S: Training Guardrail Models on Compressed Multi-turn Conversations
par: Kim, Hyunjun
Publié: (2026)

Broaden your SCOPE! Efficient Multi-turn Conversation Planning for LLMs with Semantic Space
par: Chen, Zhiliang, et autres
Publié: (2025)

REVEAL: Multi-turn Evaluation of Image-Input Harms for Vision LLM
par: Jindal, Madhur, et autres
Publié: (2025)

Krul: Efficient State Restoration for Multi-turn Conversations with Dynamic Cross-layer KV Sharing
par: Wen, Junyi, et autres
Publié: (2025)

SOMA: Efficient Multi-turn LLM Serving via Small Language Model
par: Cheng, Xueqi, et autres
Publié: (2026)

A Survey on Recent Advances in LLM-Based Multi-turn Dialogue Systems
par: Yi, Zihao, et autres
Publié: (2024)

Rethinking Patient Education as Multi-turn Multi-modal Interaction
par: Yao, Zonghai, et autres
Publié: (2026)

M2S: Multi-turn to Single-turn jailbreak in Red Teaming for LLMs
par: Ha, Junwoo, et autres
Publié: (2025)

A Decade-Scale Benchmark Evaluating LLMs' Clinical Practice Guidelines Detection and Adherence in Multi-turn Conversations
par: Tan, Andong, et autres
Publié: (2026)

Learning to Clarify: Multi-turn Conversations with Action-Based Contrastive Self-Training
par: Chen, Maximillian, et autres
Publié: (2024)

Can a Single Model Master Both Multi-turn Conversations and Tool Use? CoALM: A Unified Conversational Agentic Language Model
par: Acikgoz, Emre Can, et autres
Publié: (2025)

MINT: Evaluating LLMs in Multi-turn Interaction with Tools and Language Feedback
par: Wang, Xingyao, et autres
Publié: (2023)

X-Teaming Evolutionary M2S: Automated Discovery of Multi-turn to Single-turn Jailbreak Templates
par: Kim, Hyunjun, et autres
Publié: (2025)

AgentBoard: An Analytical Evaluation Board of Multi-turn LLM Agents
par: Ma, Chang, et autres
Publié: (2024)

Joint Detection of Fraud and Concept Drift inOnline Conversations with LLM-Assisted Judgment
par: Senol, Ali, et autres
Publié: (2025)

The Slow Drift of Support: Boundary Failures in Multi-Turn Mental Health LLM Dialogues
par: Cheng, Youyou, et autres
Publié: (2026)

Self-Anchoring Calibration Drift in Large Language Models: How Multi-Turn Conversations Reshape Model Confidence
par: Harshavardhan
Publié: (2026)

Mass-Scale Analysis of In-the-Wild Conversations Reveals Complexity Bounds on LLM Jailbreaking
par: Creo, Aldan, et autres
Publié: (2025)

Mitigating Lost in Multi-turn Conversation via Curriculum RL with Verifiable Accuracy and Abstention Rewards
par: Li, Ming, et autres
Publié: (2025)

Data Selection for Multi-turn Dialogue Instruction Tuning
par: Li, Bo, et autres
Publié: (2026)

Foot-In-The-Door: A Multi-turn Jailbreak for LLMs
par: Weng, Zixuan, et autres
Publié: (2025)

Drift-Bench: Diagnosing Cooperative Breakdowns in LLM Agents under Input Faults via Multi-Turn Interaction
par: Bao, Han, et autres
Publié: (2026)

Skill-SD: Skill-Conditioned Self-Distillation for Multi-turn LLM Agents
par: Wang, Hao, et autres
Publié: (2026)

From Self-Evolving Synthetic Data to Verifiable-Reward RL: Post-Training Multi-turn Interactive Tool-Using Agents
par: Gao, Jiaxuan, et autres
Publié: (2026)

Improving Multi-turn Dialogue Consistency with Self-Recall Thinking
par: Pang, Renning, et autres
Publié: (2026)

SafeMT: Multi-turn Safety for Multimodal Language Models
par: Zhu, Han, et autres
Publié: (2025)

Taking a turn for the better: Conversation redirection throughout the course of mental-health therapy
par: Nguyen, Vivian, et autres
Publié: (2024)

Securing Multi-turn Conversational Language Models From Distributed Backdoor Triggers
par: Tong, Terry, et autres
Publié: (2024)

Scaling LLM Multi-turn RL with End-to-end Summarization-based Context Management
par: Lu, Miao, et autres
Publié: (2025)

Phonetic Perturbations Reveal Tokenizer-Rooted Safety Gaps in LLMs
par: Aswal, Darpan, et autres
Publié: (2025)