Symbolic Neural Generation with Applications to Lead Discovery in Drug Design

arXiv:2510. 23379v2 Announce Type: replace-cross Abstract: We investigate a relatively under-explored class of hybrid neurosymbolic models that integrate symbolic learning with neural reasoning to construct data generators meeting formal correctness criteria.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

• llms
• benchmarks