# ConsistentChat: Building Skeleton-Guided Consistent Multi-Turn Dialogues for Large Language Models from Scratch

arXiv: https://arxiv.org/abs/2506.03558 · HTML: https://mlsanity.com/paper/2506.03558/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in Code and instruction data synthesis, one of 13 directions in Synthetic data and self-improvement for LLMs.
- Builds on 21 of its 33 references that are in the map.
- First posted on arXiv 4 Jun 2025.

## Placed in

- Synthetic data and self-improvement for LLMs / [Code and instruction data synthesis](https://mlsanity.com/synth/code-and-instruction-data-synthesis/)

## Builds on (most cited first, 21 of 21)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Self-Instruct: Aligning Language Models with Self-Generated Instructions](https://mlsanity.com/paper/2212.10560/) (2022)
- [G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment](https://mlsanity.com/paper/2303.16634/) (2023)
- [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://mlsanity.com/paper/2110.13900/) (2021)
- [LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Models](https://mlsanity.com/paper/2403.13372/) (2024)
- [Enhancing Chat Language Models by Scaling High-quality Instructional Conversations](https://mlsanity.com/paper/2305.14233/) (2023)
- [LMSYS-Chat-1M: A Large-Scale Real-World LLM Conversation Dataset](https://mlsanity.com/paper/2309.11998/) (2023)
- [OPT-IML: Scaling Language Model Instruction Meta Learning through the Lens of Generalization](https://mlsanity.com/paper/2212.12017/) (2022)
- [Learning to Speak and Act in a Fantasy Text Adventure Game](https://mlsanity.com/paper/1903.03094/) (2019)
- [Mitigating Catastrophic Forgetting in Large Language Models with Self-Synthesized Rehearsal](https://mlsanity.com/paper/2403.01244/) (2024)
- [Synthetic Data (Almost) from Scratch: Generalized Instruction Tuning for Language Models](https://mlsanity.com/paper/2402.13064/) (2024)
- [MemoChat: Tuning LLMs to Use Memos for Consistent Long-Range Open-Domain Conversation](https://mlsanity.com/paper/2308.08239/) (2023)
- [PPTAgent: Generating and Evaluating Presentations Beyond Text-to-Slides](https://mlsanity.com/paper/2501.03936/) (2025)
- [PersonaX: A Recommendation Agent Oriented User Modeling Framework for Long Behavior Sequence](https://mlsanity.com/paper/2503.02398/) (2025)
- [Target-oriented Proactive Dialogue Systems with Personalization: Problem Formulation and Dataset Curation](https://mlsanity.com/paper/2310.07397/) (2023)
- [Am I Me or You? State-of-the-Art Dialogue Models Cannot Maintain an Identity](https://mlsanity.com/paper/2112.05843/) (2021)
- [Instruct Once, Chat Consistently in Multiple Rounds: An Efficient Tuning Framework for Dialogue](https://mlsanity.com/paper/2402.06967/) (2024)
- [Steering Output Style and Topic in Neural Response Generation](https://mlsanity.com/paper/1709.03010/) (2017)
- [REInstruct: Building Instruction Data from Unlabeled Corpus](https://mlsanity.com/paper/2408.10663/) (2024)

## Built on by (most cited first, 5 of 5)

- [MemSearcher: Training LLMs to Reason, Search and Manage Memory via End-to-End Reinforcement Learning](https://mlsanity.com/paper/2511.02805/) (2025)
- [Not All Features Deserve Attention: Graph-Guided Dependency Learning for Tabular Data Generation with Language Models](https://mlsanity.com/paper/2507.18504/) (2025)
- [Attributes as Textual Genes: Leveraging LLMs as Genetic Algorithm Simulators for Conditional Synthetic Data Generation](https://mlsanity.com/paper/2509.02040/) (2025)
- [GraphIF: Enhancing Multi-Turn Instruction Following for Large Language Models with Relation Graph Prompt](https://mlsanity.com/paper/2511.10051/) (2025)
- [PSY-STEP: Structuring Therapeutic Targets and Action Sequences for Proactive Counseling Dialogue Systems](https://mlsanity.com/paper/2604.04448/) (2026)

## Abstract (arXiv)

Current instruction data synthesis methods primarily focus on single-turn instructions and often neglect cross-turn coherence, resulting in context drift and reduced task completion rates in extended conversations. To address this limitation, we propose Skeleton-Guided Multi-Turn Dialogue Generation, a framework that constrains multi-turn instruction synthesis by explicitly modeling human conversational intent. It operates in two stages: (1) Intent Modeling, which captures the global structure of human dialogues by assigning each conversation to one of nine well-defined intent trajectories, ensuring a coherent and goal-oriented information flow; and (2) Skeleton Generation, which constructs a structurally grounded sequence of user queries aligned with the modeled intent, thereby serving as a scaffold that constrains and guides the downstream instruction synthesis process. Based on this process, we construct ConsistentChat, a multi-turn instruction dataset with approximately 15,000 multi-turn conversations and 224,392 utterances. Experiments on the Light, Topdial, and MT-Eval benchmarks show that models fine-tuned on ConsistentChat achieve a 20-30% improvement in chat consistency and up to a 15% increase in task success rate, significantly outperforming models trained on existing single-turn and multi-turn instruction datasets.