NormGym benchmark shows frontier agents can't infer unspoken community norms
xuanalogue · x · 2026-11-20
Researchers released NormGym, an interactive benchmark for evaluating agents' normative competency: agents are placed in a virtual community with unknown norms and must learn and adapt to them through social interaction.
Key findings:
- Frontier models fail to reliably understand community norms, leaving a large gap versus humans;
- They rely almost entirely on punishment for their own violations to learn, and struggle to infer underlying norms from others' behavior;
- The authors argue building agents that act as community members — not just goal pursuers — remains an open challenge.
Paper and project page are public.
More from coding & agent
- QuixiInfer author: tell Claude 'defeat is not an option' to make it write kernels for your GPU — QuixiAI · 2026-11-20
- Dev argues terminal/tmux is the wrong form factor for managing 5-10 AI agents — DuaneJRich · 2026-11-20
- Volcano launches Agent Cloud: self-healing deploy platform built for coding agents — rohanpaul_ai · 2026-11-20
- Dusk bot: open-source Grok Bot alternative reuses your ChatGPT/Claude/Grok subscriptions, 476 GitHub stars — EXM7777 · 2026-11-20
- Theo's Restaurant Analogy Nails Agentic UX: The Whole Kitchen Waits on Your Fries Decision — seatedro · 2026-11-20
- An Interactive Cat's-Eye World in a Single HTML File — No Images, Models, or Libraries — NathanWilbanks_ · 2026-11-20