DUMA-Bench

What is DUMA-Bench?

Evaluates LLM agent security under dual-control interaction where both agent and user influence shared environment state.

Released

2026-09-21

Evaluates

Agents, Agents & Tool Use, General AI

Openness

unknown

Importer

Claire Radar

Review status

ai-name-audit-deferred

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2609.24662

DUMA-Bench paper

• Paper https://arxiv.org/abs/2609.24662

Which sources cite DUMA-Bench?

• DUMA-Bench

Related Agents, Agents & Tool Use, General AI benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 2d-fet-bench
• 3DCodeBench
• 5-Dialects-BN
• A Real-World CAD Segmentation Benchmark
• A3S-Bench
• AARRI-Bench
• ABACUS