The Benefit of Being Non-Lazy in Probabilistic {\lambda}-calculus

We consider the probabilistic applicative bisimilarity (PAB), a coinductive relation comparing the applicative behaviour of probabilistic untyped lambda terms according to a specific operational semantics. This notion has been studied with respect to the two standard parameter passing policies, call-by-value (cbv) and call-by-name (cbn), using a lazy reduction strategy not reducing within the body of a function. In particular, PAB has been proven to be fully abstract with respect to the contextual equivalence in cbv but not in lazy cbn. We overcome this issue of cbn by relaxing the laziness constraint: we prove that PAB is fully abstract with respect to the standard head reduction contextual equivalence. Our proof is based on the Leventis Separation Theorem, using probabilistic Nakajima trees as a tree-like representation of the contextual equivalence classes. Finally, we prove also that the inequality full abstraction fails, showing that the probabilistic applicative similarity is strictly contained in the contextual preorder.

[1]  C.-H. Luke Ong,et al.  Full Abstraction in the Lazy Lambda Calculus , 1993, Inf. Comput..

[2]  Michele Pagani,et al.  The Discriminating Power of the Let-In Operator in the Lazy Call-by-Name Probabilistic lambda-Calculus , 2019, FSCD.

[3]  Thomas Ehrhard,et al.  Probabilistic coherence spaces as a model of higher-order probabilistic computation , 2011, Inf. Comput..

[4]  Joël Ouaknine,et al.  Domain theory, testing and simulation for labelled Markov processes , 2005, Theor. Comput. Sci..

[5]  Hugo Paquet,et al.  Fully Abstract Models of the Probabilistic lambda-calculus , 2018, CSL.

[6]  Ugo Dal Lago,et al.  On coinductive equivalences for higher-order probabilistic functional programs , 2013, POPL.

[7]  Ugo Dal Lago,et al.  A lambda-calculus foundation for universal probabilistic programming , 2015, ICFP.

[8]  Peter Sestoft,et al.  Demonstrating Lambda Calculus Reduction , 2001, MFPS.

[9]  Ugo Dal Lago,et al.  On Applicative Similarity, Sequentiality, and Full Abstraction , 2015, Correct System Design.

[10]  Thomas Ehrhard,et al.  Probabilistic coherence spaces are fully abstract for probabilistic PCF , 2014, POPL.

[11]  Thomas Ehrhard,et al.  The Computational Meaning of Probabilistic Coherence Spaces , 2011, 2011 IEEE 26th Annual Symposium on Logic in Computer Science.

[12]  Chris Hankin,et al.  Probabilistic λ-calculus and Quantitative Program Analysis , 2004 .

[13]  Thomas Ehrhard,et al.  Full Abstraction for Probabilistic PCF , 2015, J. ACM.

[14]  Michele Pagani,et al.  Strong Adequacy and Untyped Full-Abstraction for Probabilistic Coherence Spaces , 2019, FoSSaCS.

[15]  Kim G. Larsen,et al.  Bisimulation through Probabilistic Testing , 1991, Inf. Comput..

[16]  Ugo Dal Lago,et al.  Probabilistic operational semantics for the lambda calculus , 2011, RAIRO Theor. Informatics Appl..

[17]  S. Abramsky The lazy lambda calculus , 1990 .

[18]  Giulio Manzonetto,et al.  Weighted Relational Models of Typed Lambda-Calculi , 2013, 2013 28th Annual ACM/IEEE Symposium on Logic in Computer Science.

[19]  G.D. Plotkin,et al.  LCF Considered as a Programming Language , 1977, Theor. Comput. Sci..

[20]  Thomas Leventis Probabilistic Böhm Trees and Probabilistic Separation , 2018, LICS.

[21]  Douglas J. Howe Proving Congruence of Bisimulation in Functional Programming Languages , 1996, Inf. Comput..

[22]  Ugo Dal Lago,et al.  On Probabilistic Applicative Bisimulation and Call-by-Value λ-Calculi , 2014, ESOP.