arXiv:2611.18067v1 Announce Type: new Abstract: Artificial intelligence has progressed from task-specific systems toward general models spanning perception, generation, reasoning, and autonomous interaction, making reliable evaluation essential for comparison, deployment, and post-training feedback. At scale, modern evaluation pipelines increasingly rely on LLM-as-a-Judge to produce scores, preferences, and critiques; in its fixed-context setting, the judge relies on the supplied evaluand and evaluation materials. Yet a reliable verdict may require evidence beyond the supplied evaluand and context---either because critical environmental states remain unrecorded in the static execution log, or because controversial factual claims require external verification. Passive judges cannot obtain that missing evidence before deciding, especially when evaluating agent trajectories, executable artifacts, or domain-specific reports. Recent work therefore develops active evaluators that seek missing evidence and update their grounds for judgment before deciding. We bring these advances together through Agent-as-a-Judge, focusing on how evaluators construct and control checking processes within the broader LLM-based judging landscape. We define the scope through evaluation-directed checks and the use of their results, covering both structured workflows and observation-driven control. We organize the design space by multi-agent collaboration, tool integration, memory, and planning, and show how these mechanisms supply the checks and evidence on which a verdict rests. We consolidate published scores and reproduce missing entries for key agentic judges; the comparison reveals uneven agentic gains and motivates dedicated agentic evaluation. We conclude by characterizing application demands and open challenges in cost, reliability, safety, and governance that condition trusted deployment of agentic judges at scale.
Agent-as-a-Judge
A new arXiv paper (2611.18067v1) proposes "Agent-as-a-Judge," a framework in which LLM-based evaluators actively seek missing evidence and update their grounds for judgment before issuing a verdict, rather than relying only on the supplied evaluand and static context. The paper organizes the design space around multi-agent collaboration, tool integration, memory, and planning, and consolidates published scores while reproducing missing entries for key agentic judges, finding uneven agentic gains that motivate dedicated agentic evaluation. The authors close by characterizing application demands and open challenges in cost, reliability, safety, and governance that condition trusted deployment of agentic judges at scale.
Run your AI side-project on zahid.host
EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.