Reflection AI says its Beam model was pretrained in 4 weeks on 24T high-quality tokens, giving it innate coding capabilities.
The company credits MoE stability improvements and large-scale data curation and deduplication for a base model it claims outperforms open-source base models of the same class.
It presents this strong reasoning foundation as what makes sustained reinforcement learning gains possible.
A reply · the post it answers
Sustained RL improvement was made possible by a strong foundation for reasoning.
Beam was pretrained in 4 weeks on 24T high-quality tokens to have innate coding capabilities.
Innovations in MoE stability and large-scale data curation & deduplication enabled us to produce a foundation that outperforms available open-source base models of the same class.
Source: Reflection · x.comPublished · added here
