{
  "id": 6081,
  "url": "https://arxiv.org/abs/2604.08988v3",
  "title": "SEA-Eval: A Benchmark for Evaluating Self-Evolving Agents Beyond Episodic Assessment",
  "summary": "Current LLM-based agents demonstrate strong performance in episodic task execution but remain constrained by static toolsets and episodic amnesia, failing to accumulate experience across task boundaries. This paper formalizes the Self-Evolving Agent (SEA) from the perspective of digital embodiment and continuous cross-task evolution, introduces the Evolutionary Flywheel as its minimal sufficient architecture, and presents SEA-Eval -- the first benchmark designed specifically for evaluating SEAs.",
  "authors": "Sihang Jiang, Lipeng Ma, Zhonghua Hong, Keyi Wang, Zhiyu Lu, Tengfei Wang et al.",
  "category": "research",
  "topics": "agents-autonomy",
  "orgs": null,
  "regions": null,
  "published_at": "2026-04-10T05:49:50.000Z",
  "fetched_at": "2026-07-14T16:32:15.634Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/6081",
  "original_url": "https://arxiv.org/abs/2604.08988v3",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}