AI models
ASE+ACE
Training compute
3.2×10⁸ FLOP
Parameters
324
Published
Sep 1, 1983
ASE+ACE is an AI model developed by University of Massachusetts Amherst (United States), first published in September 1983. It works in the robotics domain, on tasks such as pole balancing. It counts among the frontier models: the systems trained with the most compute of their moment.
Training it took an estimated 3.2×10⁸ FLOP of compute (estimation method: operation counting). The model has 324 parameters. It was trained on roughly 500K datapoints.
The reference paper has 4,296 citations. Epoch AI rates the confidence of this record as likely.
Full record
- Organization
- University of Massachusetts Amherst
- Country of organization
- United States
- Domain
- Robotics
- Task
- Pole balancing
- Training compute
- 3.2×10⁸ FLOP
- Compute estimation method
- Operation counting
- Parameters
- 324
- Dataset size
- 500K
- Training time
- 3 h
- Citations
- 4,296
- Epoch confidence
- Likely
More from University of Massachusetts Amherst
SourceEpoch AI, 'AI Models'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/ai-models. Licensed under CC BY 4.0.
← All ai models