Clarify Before Executing: A Self-Evolving Agent for Resolving Intent Asymmetry in 3D Tool Orchestration

2026-07-17

AI Summary

This work addresses the intent asymmetry between users’ ambiguous natural-language instructions and the precise parameter requirements of 3D modeling tools by introducing CLARE, a clarification-aware, self-evolving 3D agent. Treating intent misalignment as an opportunity for dialogue, CLARE employs a four-role decoupled cognitive architecture that autonomously generates clarification strategies through multi-turn simulated interactions, eliminating the need for handcrafted rules. Evaluated on the newly established 3D-Clarify benchmark, CLARE achieves success rates of 60.40% and 43.34% on single-step and multi-step tasks, respectively—more than doubling the performance of existing baselines. This represents the first demonstration of efficient human-AI collaborative 3D creation powered by self-evolved clarification mechanisms.

Abstract

A fundamental intent asymmetry plagues modern 3D asset creation: while state-of-the-art 3D toolchains demand precise, executable parameters, ordinary users typically provide vague, underspecified instructions. Current 3D agents treat this ambiguity as noise, defaulting to blind execution under a single-turn assumption. To address this limitation, we introduce CLARE, a clarification-aware and evolutionary 3D agent that treats intent asymmetry not as an execution error, but as an opportunity for strategic dialogue. By decoupling the generation pipeline into four specialized cognitive roles, CLARE intercepts and resolves underspecified instructions before invoking computationally expensive 3D tools to seamlessly execute tasks across five diverse domains: text-to-3D generation, single-view reconstruction, multi-view reconstruction, point cloud editing, and post-processing. Crucially, rather than relying on rigid manual rules, CLARE self-evolves its clarification policy via simulated multi-turn interactions. By optimizing a Multi-turn Reward, the agent internalizes the delicate balance between interaction efficiency and task completion. To rigorously test this, we construct 3D-Clarify, a comprehensive benchmark comprising 620 interaction scenarios with systematically injected ambiguity, missing information, and mistaken details. CLARE achieves state-of-the-art performance, with 60.40% and 43.34% success rates on single-step and multi-step tasks, respectively, more than doubling existing baselines. Both quantitative and qualitative results demonstrate that proactive clarification is the missing key to robust 3D execution. Code is available at https://github.com/xyzhu1225/CLARE.

Problem

• intent asymmetry
• 3D tool orchestration
• ambiguous instructions
• user intent clarification
• 3D asset creation

Innovation

• intent asymmetry
• self-evolving agent
• clarification-aware
• multi-turn interaction
• 3D tool orchestration