SAGEA Research Team, Basab Jha, Firoj Paudel, Ujjwal Puri, Adrian Liu, Ethan Henkel, Zhang Yuting, Mateusz Kowalczyk, Mei Huang, Choi Donghyuk, Wang Junhao
View original ↗Develop a lightweight library that implements the Inverse Reasoning (IR) verification loop for open-weights models. This would allow developers to fine-tune existing architectures to self-validate logic chains.
Suggested repo: celer-reason
"Train models that actually check their own homework before answering."
Estimated effort: 80h