Extensional Taylor Expansion
CoRR(2023)
摘要
We introduce a calculus of extensional resource terms. These are resource terms \`a la Ehrhard-Regnier, but in infinite $\eta$-long form, while retaining a finite syntax and dynamics: in particular, we prove strong confluence and normalization. Then we define an extensional version of Taylor expansion, mapping ordinary $\lambda$-terms to sets (or infinite linear combinations) of extensional resource terms: just like for ordinary Taylor expansion, the dynamics of our resource calculus allows to simulate the $\beta$-reduction of $\lambda$-terms; the extensional nature of expansion shows in that we are also able to simulate $\eta$-reduction. In a sense, extensional resource terms form a language of (non-necessarily normal) finite approximants of Nakajima trees, much like ordinary resource terms are approximants of B\"ohm-trees. Indeed, we show that the equivalence induced on $\lambda$-terms by the normalization of extensional Taylor-expansion is nothing but $H^*$, the greatest consistent sensible $\lambda$-theory. Taylor expansion has profoundly renewed the approximation theory of the $\lambda$-calculus by providing a quantitative alternative to order-based approximation techniques, such as Scott continuity and B\"ohm trees. Extensional Taylor expansion enjoys similar advantages: e.g., to exhibit models of $H^*$, it is now sufficient to provide a model of the extensional resource calculus. We apply this strategy to give a new, elementary proof of a result by Manzonetto: $H^*$ is the $\lambda$-theory induced by a well-chosen reflexive object in the relational model of the $\lambda$-calculus.
更多查看译文
关键词
expansion
AI 理解论文
溯源树
样例
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要