Sign in to view source links and access this dataset
Description
Nemotron-AIQ-Agentic-Safety-Dataset captures a range of novel safety and security contextual risks that can emerge within agentic systems. The dataset was created by NVIDIA and last updated on December 6, 2025. It is used to demonstrate the robustness of NVIDIA's open model, llama-3.3-nemotron-super-49b-v1, when deployed as a research assistant.
Use Cases
Benchmarking AI agent safety performance based on described contextual risk scenarios
Training models to identify and mitigate novel security threats in agentic systems
Researching the emergence of safety challenges within autonomous AI workflows
Strengths
Focuses on novel and contextual safety risks specific to agentic systems
Created by NVIDIA, a leading AI hardware and software company
Demonstrates the robustness of a specific, named open model (llama-3.3-nemotron-super-49b-v1)
Limitations
Description metadata is limited; actual data quality requires manual inspection after download
Column-level documentation is absent; field semantics must be inferred after download
Row count and file formats are unknown, which may limit suitability assessment
Provenance
Source
NVIDIA
Collection Method
Likely generated to test and demonstrate the safety capabilities of an AI model within the AIQ platform.
Time Range
Dataset version 1.0 released by December 2025.
Freshness
Last updated 2025-12-06 23:07:38; freshness should be verified
Geography
null
License is unknown; terms of use must be verified before application.