# You Changed Your Mind, The Model Didn't: Demystifying Intent in Multi-Turn Dialogue

arXiv: https://arxiv.org/abs/2610.06496 · HTML: https://mlsanity.com/paper/2610.06496/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 13 of its 22 references that are in the map.
- First posted on arXiv 5 Oct 2026.

## Builds on (most cited first, 13 of 13)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Program Synthesis with Large Language Models](https://mlsanity.com/paper/2108.07732/) (2021)
- [Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models](https://mlsanity.com/paper/2601.18734/) (2026)
- [On-Policy Context Distillation for Language Models](https://mlsanity.com/paper/2602.12275/) (2026)
- [LLMs Get Lost in Evolving User Intent](https://mlsanity.com/paper/2607.20734/) (2026)
- [Intent Mismatch Causes LLMs to Get Lost in Multi-Turn Conversation](https://mlsanity.com/paper/2602.07338/) (2026)
- [SEQUOR: A Multi-Turn Benchmark for Realistic Constraint Following](https://mlsanity.com/paper/2605.06353/) (2026)
- [MAIGO: Mitigating Lost-in-Conversation with History-Cleaned On-Policy Self-Distillation](https://mlsanity.com/paper/2605.27186/) (2026)
- [Trip+: Benchmarking Agents in Personalized Interactive Travel Planning](https://mlsanity.com/paper/2606.21169/) (2026)
- [Found in Conversation: LLMs Teach Themselves to Close the Multi-Turn Gap](https://mlsanity.com/paper/2605.24432/) (2026)
- [SpecPath: Testing Coding Agents Across Contract-Equivalent Specification Histories](https://mlsanity.com/paper/2608.09799/) (2026)

## Abstract (arXiv)

When a large language model handles a multi-turn task and a user proposes a change but ultimately rejects it, the model should continue as if nothing changed. We find a surprising failure: merely mentioning a rejected change can derail task execution, even when the user's final intent remains unchanged. To systematically study language model behavior under evolving user intent, we introduce Intent-Eval, a controlled benchmark spanning tool actions, code, databases, and mathematics. Across diverse tasks, models are vulnerable to both rejected proposals and superseded requirements, consistent with mentioned-as-in-effect confusion: conversational content is treated as active requirements even after it has been rejected or replaced. Accuracy degradation can deepen or persist as interaction continues, highlighting the need to distinguish what has been mentioned from what remains in effect. Building on this insight, we propose Intent-OPSD, a decision-conditioned on-policy self-distillation framework with Teacher and Student initialized from the same model. The frozen Teacher provides active-intent supervision from the complete task matching the user's decision, training the Student on the full dialogue to follow active requirements reflecting user intent.