Recent advances in large language models (LLMs) have improved Verilog generation. However, most training pipelines distill natural-language reasoning traces whose intermediate design decisions are weakly structured and are primarily selected according to the correctness of the final generated RTL. We present MCTS-SFT,...
Matthew DeLorenzo, Kevin Tieu, Prithwish Jana et al.· 0 citations
Financial Artificial Intelligence (AI) governance and model risk management (MRM) emphasize independent validation, transparency, and effective challenge of AI-generated computations. Text-to-SQL translates natural-language questions into Structured Query Language (SQL), but queries that agree on one database can still...
Piyush Jha, Cruise Song, Zhengyu Li et al.· 0 citations
Many scientific discovery problems require searching combinatorial hypothesis spaces under complex domain constraints. Reinforcement learning (RL) offers a promising approach, but existing methods rely on scalar rewards that provide limited information about why candidate solutions fail, leading agents to repeatedly ex...
Piyush Jha, Jake Rudolph, V. Knapp-Pérez et al.· 3 citations
CDES is introduced, which extends evolutionary search with enforceable restrictions derived from failed candidates, recorded as certificates of assumptions, checker evidence, and justified restrictions, to CPU-to-GPU translation of two particle-simulation functions from the Geant4 toolkit.
Piyush Jha, A. Ghosh, Vijay Ganesh· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.