paper-with-me

홈 › Papers

Seed2Scale: A Self-Evolving Data Engine for Embodied AI via Small to Large Model Synergy and Multimodal Evaluation

2026-03-09 · Cong Tai, Zhaoyu Zheng, Haixu Long, Hansheng Wu, Zhengbin Long, Haodong Xiang, Rong Shi, Zhuo Cui, Shizhuang Zhang, Gang Qiu, He Wang, Ruifeng Li, Biao Liu, Zhenzhe Sun, Tao Shen arxiv

Existing data generation methods suffer from exploration limits, embodiment gaps, and low signal-to-noise ratios, leading to performance degradation during self-iteration. To address these challenges, we propose Seed2Scale, a self-evolving data engine that overcomes the data bottleneck through a heterogeneous synergy of "small-model collection, large-model evaluation, and target-model learning". Starting with as few as four seed demonstrations, the engine employs the lightweight Vision-Language-Action model, SuperTiny, as a dedicated collector, leveraging its strong inductive bias for robust exploration in parallel environments. Concurrently, a pre-trained Vision-Language Model is integrated as a Verifer to autonomously perform success/failure judgment and quality scoring for the massive generated trajectories. Seed2Scale effectively mitigates model collapse, ensuring the stability of the self-evolution process. Experimental results demonstrate that Seed2Scale exhibits signifcant scaling potential: as iterations progress, the success rate of the target model shows a robust upward trend, achieving a performance improvement of 131.2%. Furthermore, Seed2Scale signifcantly outperforms existing data augmentation methods, providing a scalable and cost-effective pathway for the large-scale development of Generalist Embodied AI. Project page: https://terminators2025.github.io/Seed2Scale.github.io

📄 PDF Abstract BibTeX arXiv:2603.08260

Code (0)

등록된 구현이 없습니다.

Tasks

Data Augmentation

Similar Papers 제목 키워드 기반

SeedPolicy: Horizon Scaling via Self-Evolving Diffusion Policy for Robot Manipulation

2026-03-05 · Youqiang Gui, Yuxuan Zhou, Shen Cheng, Xinyang Yuan 외 arxiv

Imitation Learning (IL) enables robots to acquire manipulation skills from expert demonstrations. Diffusion Policy (DP) models multi-modal expert behaviors but degrades when naively increasing stacked observation horizon…

Robot Manipulation

SEED: Self-Evolving On-Policy Distillation for Agentic Reinforcement Learning

2026-07-16 · Jinyang Wu, Shuo Yang, Zhengxi Lu, Fan Zhang 외 arxiv

Large language models are increasingly trained as interactive agents for long-horizon tasks involving multi-turn interaction, tool use, and environment feedback. Outcome-based reinforcement learning (RL) provides a pract…

Reinforcement LearningDecision Making

Middo: Model-Informed Dynamic Data Optimization for Enhanced LLM Fine-Tuning via Closed-Loop Learning

2025-08-29 · Zinan Tang, Xin Gao, Qizhi Pei, Zhuoshi Pan 외 arxiv

Supervised Fine-Tuning (SFT) Large Language Models (LLM) fundamentally rely on high-quality training data. While data selection and data synthesis are two common strategies to improve data quality, existing approaches of…

GrowLoop: Self-Evolving Conversation Evaluation Seeded by Human

2026-05-26 · Yihang Lin, Yunze Gao, Zeyang Lin, Dongbo Li 외 arxiv

With the rapid advancement of large language models, evaluating human-likeness in open-ended conversation has become increasingly important. However, human-likeness is a form of tacit knowledge that humans perceive intui…

Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses

2026-04-28 · Jiahang Lin, Shichun Liu, Chengjun Pan, Lizhi Lin 외 arxiv

Harnesses are now central to coding-agent performance, mediating how models interact with tools and execution environments. Yet harness engineering remains a manual craft, because automating it faces a heterogeneous acti…