synthetic-benchmarks

Here are 4 public repositories matching this topic...

tombudd / agent-action-audit-template

Clean-room public template for synthetic agent action audit records and human review checkpoints.

json-schema governance ai-safety human-in-the-loop audit-trail software-agents red-teaming responsible-ai ai-evaluation ai-governance human-ai-collaboration agentic-ai auditability ai-audit approval-gates synthetic-benchmarks action-review

Updated Jul 7, 2026
Python

tombudd / human-ai-governance-lab

Star

Clean-room project exploring reviewable, testable, accountable human-AI collaboration workflows.

workflow governance receipts ai-safety human-in-the-loop software-agents red-teaming responsible-ai risk-classification ai-evaluation ai-governance human-ai-collaboration agentic-ai auditability ai-audit approval-gates synthetic-benchmarks

Updated Jul 6, 2026
Python

atilavahedian / forgetnet

Star

Pure-PyTorch plastic-memory sequence model: local attention plus bounded differentiable memory, surprise-gated writes, forgetting, synthetic benchmarks, and ablations.

machine-learning transformers pytorch attention ai-research sequence-modeling long-context memory-augmented-networks differentiable-memory synthetic-benchmarks

Updated Jul 28, 2026
Python

nshkrdotcom / superposition_zoo

Sponsor

Star

A research harness for studying how sequence-mixing architectures (softmax attention, linear attention, state-space models, delta-rule associative memory, and more) causally shape superposition and feature isolation in learned representations, using ground-truth synthetic benchmarks instead of natural-language probing.

Updated Jul 5, 2026
Python

Add this topic to your repo

To associate your repository with the synthetic-benchmarks topic, visit your repo's landing page and select "manage topics."

Learn more