UpBench: A Dynamically Evolving Real-World Labor-Market Agentic Benchmark Framework Built for Human-Centric AI

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yi, Darvin, Liu, Teng, Terzolo, Mattie, Hasson, Lance, Sinha, Ayan, Mendes, Pablo, Rabinovich, Andrew
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915671332356096
author Yi, Darvin
Liu, Teng
Terzolo, Mattie
Hasson, Lance
Sinha, Ayan
Mendes, Pablo
Rabinovich, Andrew
author_facet Yi, Darvin
Liu, Teng
Terzolo, Mattie
Hasson, Lance
Sinha, Ayan
Mendes, Pablo
Rabinovich, Andrew
contents As large language model (LLM) agents increasingly undertake digital work, reliable frameworks are needed to evaluate their real-world competence, adaptability, and capacity for human collaboration. Existing benchmarks remain largely static, synthetic, or domain-limited, providing limited insight into how agents perform in dynamic, economically meaningful environments. We introduce UpBench, a dynamically evolving benchmark grounded in real jobs drawn from the global Upwork labor marketplace. Each task corresponds to a verified client transaction, anchoring evaluation in genuine work activity and financial outcomes. UpBench employs a rubric-based evaluation framework, in which expert freelancers decompose each job into detailed, verifiable acceptance criteria and assess AI submissions with per-criterion feedback. This structure enables fine-grained analysis of model strengths, weaknesses, and instruction-following fidelity beyond binary pass/fail metrics. Human expertise is integrated throughout the data pipeline (from job curation and rubric construction to evaluation) ensuring fidelity to real professional standards and supporting research on human-AI collaboration. By regularly refreshing tasks to reflect the evolving nature of online work, UpBench provides a scalable, human-centered foundation for evaluating agentic systems in authentic labor-market contexts, offering a path toward a collaborative framework, where AI amplifies human capability through partnership rather than replacement.
format Preprint
id arxiv_https___arxiv_org_abs_2511_12306
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle UpBench: A Dynamically Evolving Real-World Labor-Market Agentic Benchmark Framework Built for Human-Centric AI
Yi, Darvin
Liu, Teng
Terzolo, Mattie
Hasson, Lance
Sinha, Ayan
Mendes, Pablo
Rabinovich, Andrew
Artificial Intelligence
Computers and Society
As large language model (LLM) agents increasingly undertake digital work, reliable frameworks are needed to evaluate their real-world competence, adaptability, and capacity for human collaboration. Existing benchmarks remain largely static, synthetic, or domain-limited, providing limited insight into how agents perform in dynamic, economically meaningful environments. We introduce UpBench, a dynamically evolving benchmark grounded in real jobs drawn from the global Upwork labor marketplace. Each task corresponds to a verified client transaction, anchoring evaluation in genuine work activity and financial outcomes. UpBench employs a rubric-based evaluation framework, in which expert freelancers decompose each job into detailed, verifiable acceptance criteria and assess AI submissions with per-criterion feedback. This structure enables fine-grained analysis of model strengths, weaknesses, and instruction-following fidelity beyond binary pass/fail metrics. Human expertise is integrated throughout the data pipeline (from job curation and rubric construction to evaluation) ensuring fidelity to real professional standards and supporting research on human-AI collaboration. By regularly refreshing tasks to reflect the evolving nature of online work, UpBench provides a scalable, human-centered foundation for evaluating agentic systems in authentic labor-market contexts, offering a path toward a collaborative framework, where AI amplifies human capability through partnership rather than replacement.
title UpBench: A Dynamically Evolving Real-World Labor-Market Agentic Benchmark Framework Built for Human-Centric AI
topic Artificial Intelligence
Computers and Society
url https://arxiv.org/abs/2511.12306