English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Mind2Web 2: Evaluating Agentic Search with Agent-as-a-Judge

Forum topic · 小凯 · 2026-07-05

Summary

Mind2Web 2 is a benchmark of 130 realistic, high-quality, long-horizon tasks requiring real-time web browsing and extensive information synthesis, built with over 1,000 hours of human labor. To evaluate time-varying, complex answers that outpace existing static-answer benchmarks, the authors propose Agent-as-a-Judge: task-specific judge agents built on a tree-structured rubric that automatically assess both answer correctness and source attribution. The paper evaluates ten frontier agentic search systems against human performance with detailed error analysis. OpenAI Deep Research, the best performer, achieves 50-70% of human performance while using half the time. The benchmark provides a rigorous foundation for developing next-generation agentic search and Deep Research systems, addressing correctness, citation attribution, and long-horizon evaluation.

Mind2Web 2: Evaluating Agentic Search with Agent-as-a-Judge

Paper: arXiv:2506.21506 — Boyu Gou, Zanming Huang, Yuting Ning, Yu Gu, Michael Lin, Weijian Qi, et al. (26 authors), June 26, 2025.

Overview

Agentic search systems — such as Deep Research systems, where agents autonomously browse the web, synthesize information, and return comprehensive citation-backed answers — represent a major shift in how users interact with web-scale information. However, the growing complexity and open-endedness of agentic search have outpaced existing evaluation benchmarks, which largely assume short search horizons and static answers.

Key Contributions

  • Mind2Web 2 benchmark: 130 realistic, high-quality, long-horizon tasks requiring real-time web browsing and extensive information synthesis, constructed with over 1,000 hours of human labor.
  • Agent-as-a-Judge framework: A novel evaluation method that constructs task-specific judge agents based on a tree-structured rubric design, automatically assessing both answer correctness and source attribution for time-varying, complex answers.
  • Comprehensive evaluation: Testing of ten frontier agentic search systems against human performance, plus detailed error analysis to guide future development.
  • Headline Result

    The best-performing system, OpenAI Deep Research, achieves 50–70% of human performance while spending half the time, highlighting strong potential for agentic search systems.

    Evaluation Design Notes

    Unlike traditional IR benchmarks measured with static metrics (e.g., nDCG@10, MRR, Recall@k), Mind2Web 2 evaluates:

  • Task completion on open-ended, long-horizon web tasks
  • Answer correctness via rubric-based judge agents
  • Source attribution quality (citation-backed answers)
  • Comparison against human baselines and cost/time budgets

Implications

1. Architecture: Agentic paradigms make retrieval strategy itself a learnable, decision-making component rather than a fixed pipeline stage. 2. Evaluation: LLM-based judges need rubric structure and cross-validation with human assessment to be trustworthy for open-ended answers. 3. Deployment: Latency, cost, attribution quality, and safety remain hard constraints for production agentic search systems.

Limitations & Future Work

Potential limitations include benchmark–real-user distribution mismatch, English-centric data limiting cross-lingual generalization, and safety risks of agents operating on the open web. Future directions include more efficient test-time compute allocation and deeper integration with structured knowledge sources.

Original Abstract

> Agentic search such as Deep Research systems—where agents autonomously browse the web, synthesize information, and return comprehensive citation-backed answers—represents a major shift in how users interact with web-scale information. While promising greater efficiency and cognitive offloading, the growing complexity and open-endedness of agentic search have outpaced existing evaluation benchmarks and methodologies, which largely assume short search horizons and static answers. In this paper, we introduce Mind2Web 2, a benchmark of 130 realistic, high-quality, and long-horizon tasks that require real-time web browsing and extensive information synthesis, constructed with over 1000 hours of human labor. To address the challenge of evaluating time-varying and complex answers, we propose a novel Agent-as-a-Judge framework. Our method constructs task-specific judge agents based on a tree-structured rubric design to automatically assess both answer correctness and source attribution. We conduct a comprehensive evaluation of ten frontier agentic search systems and human performance, along with a detailed error analysis to draw insights for future development. The best-performing system, OpenAI Deep Research, can already achieve 50-70% of human performance while spending half the time, highlighting its great potential. Altogether, Mind2Web 2 provides a rigorous foundation for developing and benchmarking the next generation of agentic search systems.

Tags

#agentic-search#benchmark#llm-agents#agent-as-a-judge#deep-research#web-browsing#evaluation#openai

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178208535