1,874
Python
Inspect: A framework for large language model evaluations
Last pushed: 2 April 2026License: MIT License
github.com/UKGovernmentBEISDepartment for Science, Innovation and Technology (DSIT)
Ministerial department
https://www.gov.uk/government/organisations/department-for-science-innovation-and-technology
Total FTE: 2,275·Digital & data FTE: 130
Sub-organisations: Government Digital Service (GDS), Intellectual Property Office (IPO), Ordnance Survey (OS), UK Space Agency, UK Research and Innovation (UKRI), Met Office
Stars of active repositories
2,653
Active repositories
45
Total repositories
148
GitHub accounts
Repositories
Showing 1–10 of 45 repositories, sorted by stars
419
Python
Collection of evals for Inspect AI
Last pushed: 2 April 2026License: MIT License
github.com/UKGovernmentBEIS174
Python
ControlArena is a collection of settings, model organisms and protocols - for running control experiments.
Last pushed: 31 March 2026License: MIT License
github.com/UKGovernmentBEIS48
Python
Last pushed: 19 March 2026License: MIT License
github.com/UKGovernmentBEIS29
Python
A Kubernetes sandbox environment for use with inspect_ai
Last pushed: 23 March 2026License: MIT License
github.com/UKGovernmentBEIS25
Python
An Inspect extension for agentic cyber evaluations
Last pushed: 24 February 2026License: MIT License
github.com/UKGovernmentBEIS7
Python
Last pushed: 29 March 2026License: MIT License
github.com/UKGovernmentBEIS7
HTML
Accompanying code for Async Control: Stress-testing Asynchronous Control Measures for LLM Agents paper
Last pushed: 25 March 2026License: MIT License
github.com/UKGovernmentBEIS6
Report Official Development Assistance (RODA)
Last pushed: 27 March 2026License: MIT License
github.com/UKGovernmentBEIS5
Reproducing "Natural Emergent Misalignment from Reward Hacking" (MacDiarmid et al., Anthropic 2025) with open-source models. Includes reward-hackable RL environments, misalignment evaluations, training configs, and evaluation scripts. Models trained on OLMo (7B, 32B) and GPT-OSS (20B, 120B).
Last pushed: 30 March 2026License: MIT License
github.com/UKGovernmentBEISPage 1 of 5