research

We revisit Tarski's Algebra of Relations (TAR), an old formalism of relations predating Codd's relational algebra by over 100 years, as a new foundation for relational databases. We argue TAR provides a better abstraction at both the semantic level and the physical level, in the context of modern application code and system architecture. To demonstrate the strengths of TAR, we design and implement Prela, a compositional and controllable query language, and show that queries written in Prela are concise, clear, and efficient.

Large Language Models (LLMs) are increasingly used to evaluate the reliability of news content, yet the mechanisms driving these judgments remain poorly understood. In this work, we investigate whether LLMs assess news trustworthiness based on substantive journalistic quality or superficial linguistic signals. We deploy a multi-stage pipeline combining an LLM-as-a-Judge classification baseline, causal feature perturbation via controlled counterfactual rewrites, and mechanistic attribution via Integrated Gradients. We find that models rely heavily on institutional-register tokens as proxies for trustworthiness, a shallow heuristic that mirrors wire-service style rather than factual accuracy.