Call for Papers @RLEval Workshop ACM CAIS 2026

[Aziza Mirsaidova]

We're organizing RLEval: Methods and Reinforcement Learning Environments for Evaluating AI Agents workshop at the ACM Conference on AI and Agentic Systems (ACM CAIS), and we'd love your submission.

We invite 4-page short papers on evaluation methods, RL environment design, benchmarks, and real-world case studies for AI agents. RLEval aims to bring together researchers and practitioners committed to rigorous, systematic evaluation of agentic systems, with a particular focus on reinforcement learning environments as a foundation for reproducible agent measurement.

Topics of interest include, but are not limited to:

• Evaluation methods: interventional evaluation, causal and counterfactual approaches, automated graders (verifiers, rubrics, LLM-as-judge)
• Data and benchmarks: new benchmarks, analyses of existing benchmarks, benchmark design and validity
• RL environments: what makes an effective environment, synthetic data, tool design, software frameworks
• Enterprise case studies: evaluation in production, environment design for specific use cases, lessons from deployments
• Agent capability evaluation: code execution, NL2SQL, computer use, multimodal I/O, MCP tools, skills, memory, web search

Key dates:
Submission deadline: May 20, 2026 (AoE)

Workshop website: https://rl-eval.github.io/
Conference website: https://www.caisconf.org/
Submission site: https://lnkd.in/geUa-nqb

Looking forward for your submission!