Clarification Is Not Correction: LLMs Fail to Let Go

arXiv:2609.25337 · cs.AI · Submitted 2026-09-21 · Read on arXiv

cs.AI

Submitted: 2026-09-21

Updated: 2026-09-21

Comments: 17 pages

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Trace — 13× in this paper · explained in 2 episodes
• Benchmark — 10× in this paper · explained in 2 episodes
• Robustness — 7× in this paper · explained in 5 episodes
• State Management — 5× in this paper · explained in 3 episodes
• Ablation — 3× in this paper · explained in 1 episode
• Benchmarking — 2× in this paper · explained in 3 episodes
• Consistency — 2× in this paper · explained in 3 episodes
• Reasoning Traces — 2× in this paper · explained in 3 episodes
• Self-consistency — 2× in this paper · explained in 2 episodes
• Chain-of-Thought Prompting — 1× in this paper · explained in 2 episodes
• Core — 1× in this paper · explained in 2 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes

Sources

• LLMs Get Lost In Multi-Turn Conversation
• MTR-Bench: A Comprehensive Benchmark for Multi-Turn Reasoning Evaluation
• TurnBench-MS: A Benchmark for Evaluating Multi-Turn, Multi-Step Reasoning in Large Language Models
• ClarifyMT-Bench: Benchmarking and Improving Multi-Turn Clarification for Conversational Large Language Models
• When and What to Ask: AskBench and Rubric-Guided RLVR for LLM Clarification
• ClarEval: A Benchmark for Evaluating Clarification Skills of Code Agents under Ambiguous Instructions
• CLARITY: A Framework and Benchmark for Conversational Language Ambiguity and Unanswerability in Interactive NL2SQL Systems
• Structured Uncertainty guided Clarification for LLM Agents
• MemGPT: Towards LLMs as Operating Systems
• LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems
• VSAL: A Vision Solver with Adaptive Layouts for Graph Property Detection