Dr. Zero: Self-Evolving Search Agents without Training Data

Paper: https://arxiv.org/abs/2601.07055 Authors: Zhenrui Yue, Kartikeya Upasani, Xianjun Yang, Suyu Ge, Shaoliang Nie, Yuning Mao, et al. (8 authors total) Published: 2026-01-11 Topic area: Agentic Search

Overview

As high-quality training data becomes increasingly difficult to obtain, data-free self-evolution has emerged as a promising paradigm: LLMs autonomously generate and solve complex problems to improve their own reasoning. However, multi-turn search agents struggle with this approach due to limited question diversity and the substantial compute required for multi-step reasoning and tool use.

Key Contributions

• Dr. Zero framework: enables search agents to self-evolve with no training data at all.
• Self-evolution feedback loop: a *proposer* generates diverse questions to train a *solver* initialized from the same base model. As the solver improves, it incentivizes the proposer to produce increasingly difficult yet still-solvable tasks — an automated curriculum refining both agents.
• HRPO (hop-grouped relative policy optimization): clusters structurally similar questions to construct group-level baselines, minimizing the sampling overhead of evaluating each query's individual difficulty and solvability. This significantly reduces solver training compute without compromising performance or stability.

Results

Extensive experiments show that the fully data-free Dr. Zero matches or surpasses fully supervised search agents, demonstrating that complex reasoning and search capabilities can emerge solely through self-evolution.

Original Abstract (verbatim)

> As high-quality data becomes increasingly difficult to obtain, data-free self-evolution has emerged as a promising paradigm. This approach allows large language models (LLMs) to autonomously generate and solve complex problems, thereby improving their reasoning capabilities. However, multi-turn search agents struggle in data-free self-evolution due to the limited question diversity and the substantial compute required for multi-step reasoning and tool using. In this work, we introduce Dr. Zero, a framework enabling search agents to effectively self-evolve without any training data. In particular, we design a self-evolution feedback loop where a proposer generates diverse questions to train a solver initialized from the same base model. As the solver evolves, it incentivizes the proposer to produce increasingly difficult yet solvable tasks, thus establishing an automated curriculum to refine both agents. To enhance training efficiency, we also introduce hop-grouped relative policy optimization (HRPO). This method clusters structurally similar questions to construct group-level baselines, effectively minimizing the sampling overhead in evaluating each query's individual difficulty and solvability. Consequently, HRPO significantly reduces the compute requirements for solver training without compromising performance or stability. Extensive experiment results demonstrate that the data-free Dr. Zero matches or surpasses fully supervised search agents, proving that complex reasoning and search capabilities can emerge solely through self-evolution.

Context and Significance

Dr. Zero sits at the intersection of agentic search and large-scale IR systems. Classical search pipelines treat retrieval, ranking, and generation separately; LLM-era agentic search instead makes the number and strategy of retrieval steps learnable decisions. Data-free self-evolution addresses a key bottleneck — the scarcity of high-quality supervised data for training multi-step search behavior — while HRPO addresses the compute cost that typically makes such training impractical.

Note: quantitative benchmark details should be verified against the original PDF; this summary is based on the abstract and public metadata.