Tag
XLOG is a CUDA-native logic programming engine designed for neurosymbolic integration, combining neural perception with symbolic reasoning through a shared typed frontend and CUDA runtime, achieving significant performance improvements in training and inference.
This paper introduces a production enterprise analytics system that inverts the interaction model from question-first to analyst-first through domain-expert skills and knowledge compilation, enabling proactive analytics with verified metrics and suggested questions.
Presents Moose, a neuro-symbolic method that compiles OWL 2 EL ontologies into Sentential Decision Diagrams for differentiable weighted model counting, enabling latent concept learning under partial supervision and providing the first reasoning-shortcut analysis in an OWL EL setting.
This paper formalizes Streaming Knowledge Compilation for LLM wikis, introducing a materiality signal to proactively pin important documents from a streaming corpus under a token budget. It proves an O(√(T log K)) regret bound and validates the approach in finance and Wikipedia domains, showing that regret analysis is a reliable evaluation metric.
The paper introduces WiCER, an iterative algorithm for compiling domain knowledge into LLM Wiki systems to minimize information loss and catastrophic failure rates during knowledge distillation. It demonstrates that this approach improves upon full-context KV cache inference by preserving critical facts better than blind compilation methods.
The paper introduces Cognitive Agent Compilation (CAC), a framework that uses teacher LLMs to compile problem-solving knowledge into explicit, inspectable agents for educational applications. It aims to address the lack of controllability and explainability in standard LLMs by separating knowledge representation from policy and verification rules.