# Recon: Reconstruction-Guided Reasoning Synthesis for User Modeling

arXiv: https://arxiv.org/abs/2605.26969 · HTML: https://mlsanity.com/paper/2605.26969/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 15 of its 31 references that are in the map.
- First posted on arXiv 26 May 2026.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 15 of 15)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [OpenAI GPT-5 System Card](https://mlsanity.com/paper/2601.03267/) (2025)
- [Personalized Soups: Personalized Large Language Model Alignment via Post-hoc Parameter Merging](https://mlsanity.com/paper/2310.11564/) (2023)
- [LLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individuals](https://mlsanity.com/paper/2411.10109/) (2024)
- [Personalized Language Modeling from Personalized Human Feedback](https://mlsanity.com/paper/2402.05133/) (2024)
- [CollabLLM: From Passive Responders to Active Collaborators](https://mlsanity.com/paper/2502.00640/) (2025)
- [Flipping the Dialogue: Training and Evaluating User Language Models](https://mlsanity.com/paper/2510.06552/) (2025)
- [HumanLM: Simulating Users with State Alignment Beats Response Imitation](https://mlsanity.com/paper/2603.03303/) (2026)
- [Can LLM Agents Simulate Multi-Turn Human Behavior? Evidence from Real Online Customer Behavior Data](https://mlsanity.com/paper/2503.20749/) (2025)
- [Aligning Language Models from User Interactions](https://mlsanity.com/paper/2603.12273/) (2026)
- [How to Train Your Advisor: Steering Black-Box LLMs with Advisor Models](https://mlsanity.com/paper/2510.02453/) (2025)
- [FSPO: Few-Shot Optimization of Synthetic Preferences Personalizes to Real Users](https://mlsanity.com/paper/2502.19312/) (2025)
- [How to Steal Reasoning Without Reasoning Traces](https://mlsanity.com/paper/2603.07267/) (2026)
- [ParaStudent: Closing the Sim2Real Gap in User Simulators for AI Tutor Evaluation](https://mlsanity.com/paper/2507.12674/) (2025)

## Abstract (arXiv)

User modeling aims to use language models (LMs) to mimic an individual's behavior from a corpus of past context-action pairs (e.g., conversation turns), enabling the simulation of users in settings like behavioral science, human-AI collaboration, and market research. Recent approaches augment these corpora with synthesized reasoning traces, typically generated by conditioning on both context and action. However, such conditioning constitutes post-hoc rationalization rather than reasoning: the trace is guaranteed to justify the action, but may not encode the underlying latent causal decision paths. We propose Recon, which uses action reconstruction to score reasoning traces by their predictive power: given a context and candidate reasoning, a reconstruction model predicts the action, and reconstruction fidelity determines reasoning quality. Across four domains, Recon achieves a 54.7% win rate over Backward Synthesis, a standard post-hoc rationalization baseline. Further, we find that training a reasoning synthesis model with rewards derived from Recon improves downstream user modeling performance, achieving a win rate of up to 70.0% over baselines. We further show that Recon-synthesized reasoning transfers across models, and improves user modeling beyond the reconstruction model. Our work demonstrates that post-hoc rationalization is insufficient for reasoning synthesis, and that useful and interpretable reasoning should naturally elicit the action from the context.