BEGIN:VCALENDAR
VERSION:2.0
PRODID:researchseminars.org
CALSCALE:GREGORIAN
X-WR-CALNAME:researchseminars.org
BEGIN:VEVENT
SUMMARY:Beniamino Accattoli (Inria\, France)
DTSTART:20210422T170000Z
DTEND:20210422T180000Z
DTSTAMP:20260404T131154Z
UID:LoReL/1
DESCRIPTION:Title: <a href="https://stable.researchseminars.org/talk/LoReL
 /1/">The Machinery of Interaction</a>\nby Beniamino Accattoli (Inria\, Fra
 nce) as part of LoReL Seminar\n\n\nAbstract\nGirard's geometry of interact
 ion suggests an unusual way of evaluating lambda terms\, the interaction a
 bstract machine (IAM)\, first studied by Mackie and Danos & Regnier in the
  nineties\, using linear logic proof nets. In particular\, the IAM does no
 t use environments\, in contrast to the mainstream approach to abstract ma
 chines. In this talk I shall give a slow paced introduction to the IAM\, u
 sing an alternative presentation based on ordinary lambda terms. If time p
 ermits\, I shall also discuss how to measure the time and space consumptio
 ns of the IAM via multi types. This talk is an overview of joint works wit
 h Ugo Dal Lago and Gabriele Vanoni.\n
LOCATION:https://stable.researchseminars.org/talk/LoReL/1/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Alberto Pardo (Universidad de la República\, Uruguay)
DTSTART:20210429T170000Z
DTEND:20210429T180000Z
DTSTAMP:20260404T131154Z
UID:LoReL/2
DESCRIPTION:Title: <a href="https://stable.researchseminars.org/talk/LoReL
 /2/">Formalización de una transformación entre programas seguros usando 
 un abordaje internalista</a>\nby Alberto Pardo (Universidad de la Repúbli
 ca\, Uruguay) as part of LoReL Seminar\n\n\nAbstract\nLos lenguajes funcio
 nales fuertemente tipados\, como Haskell\, OCaml\, Agda o Idris\, entre ot
 ros\, se destacan por sus sistemas de tipos. Gracias a sus poderosos mecan
 ismos de tipado (llamense tipos dependientes\, GADTs\, funciones a nivel d
 e tipos\, múltiples kinds\, type classes\, etc) uno puede ir más allá d
 el uso habitual que tienen los tipos en un lenguaje de programación y pas
 ar a usarlos para para representar propiedades de los datos que manipula u
 n programa (los llamados invariantes de tipos). Más allá del impacto (y 
 dificultad) que esto pueda implicar en la tarea de programación\, un beni
 ficio importante es que los invariantes pasan a formar parte del código o
 rdinario del programa y por lo tanto pueden ser verificados en tiempo de c
 ompilación por el chequeador de tipos del lenguaje. Es decir\, sólo son 
 aceptados aquellos programas que preservan los invariantes que tienen defi
 nidos en sus tipos. Esta metodología de programación juega un rol clave\
 , por ejemplo\, en el desarrollo de software confiable. Al momento de codi
 ficar los invariantes a nivel de los tipos existen distintas alternativas.
  Una es la estándar\, en la que uno asocia los invariantes a través de p
 redicados por fuera de la definición de los tipos. A este abordaje se le 
 llama "externalista". Otra opción es usando un abordaje "internalista"\, 
 en el que uno codifica el invariante como parte de la propia definición d
 el tipo. De esta forma\, se chequea la propiedad en la propia construcció
 n de los términos del tipo y por lo tanto sólo es posible construir aque
 llos términos que satisfacen el invariante. En esta charla analizaremos u
 n ejemplo en particular donde aplicamos esta metodología de programación
  con invariantes. En concreto\, vamos a mostrar la formalización en Agda 
 de una transformación entre programas seguros (de un lenguaje imperativo 
 simple) que preserva la propiedad de no interferencia. Dicha transformaci
 ón toma programas seguros según un sistema flow-sensitive (en el que el 
 nivel de seguridad de las variables puede mudar) y los transforma en progr
 amas seguros tipables según un sistema flow-insensitive (en el que el niv
 el de seguridad de las variables es fijo). Lo interesante de nuestra forma
 lización es que está desarrollada usando el abordaje internalista.\n
LOCATION:https://stable.researchseminars.org/talk/LoReL/2/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Octavio Malherbe (Universidad de la República\, Uruguay)
DTSTART:20210506T170000Z
DTEND:20210506T180000Z
DTSTAMP:20260404T131154Z
UID:LoReL/3
DESCRIPTION:Title: <a href="https://stable.researchseminars.org/talk/LoReL
 /3/">Categorías indexadas\, tripos y realizabilidad</a>\nby Octavio Malhe
 rbe (Universidad de la República\, Uruguay) as part of LoReL Seminar\n\n\
 nAbstract\nEn esta charla comenzaremos  con una introducción a las catego
 rías indexadas para luego adentrarnos con un ejemplo de estas mediante la
  noción de tripos. Esta herramienta nos permite modelar la lógica de alt
 o orden y nos va  permitir  trabajar con ciertas abstracciones de la reali
 zabilidad clásica de Krivine  denominadas Abstract Krivine Structure (AKS
 ) y ciertas álgebras combinatorias (denominadas full adjunction ordered c
 ombinatory algebras).  Veremos cómo de una AKS se determina una de estas 
 álgebras y en el otro sentido estas determinan un AKS para que sus tripos
  sean equivalentes.\n
LOCATION:https://stable.researchseminars.org/talk/LoReL/3/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Andrés Viso (Software Heritage\, France and UNQ\, Argentina)
DTSTART:20210527T170000Z
DTEND:20210527T180000Z
DTSTAMP:20260404T131154Z
UID:LoReL/4
DESCRIPTION:Title: <a href="https://stable.researchseminars.org/talk/LoReL
 /4/">Encoding Tight Typing in a Unified Framework</a>\nby Andrés Viso (So
 ftware Heritage\, France and UNQ\, Argentina) as part of LoReL Seminar\n\n
 \nAbstract\nThis joint work with Delia Kesner explores how the intersectio
 n type theories of call-by-name (CBN) and call-by-value (CBV) can be unifi
 ed in a more general framework provided by call-by-push-value (CBPV). Inde
 ed\, we propose tight type systems for CBN and CBV that can be both encode
 d in a tight type system for CBPV. All such systems are quantitative\, in 
 the sense that they provide exact information about the length of normaliz
 ation sequences to normal form as well as the size of these normal forms. 
 Moreover\, the length of sequences to normal forms are discriminated accor
 ding to their multiplicative/exponential nature\, a concept inherited from
  linear logic.\n
LOCATION:https://stable.researchseminars.org/talk/LoReL/4/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Éric Tanter (Universidad de Chile\, Chile)
DTSTART:20210520T170000Z
DTEND:20210520T180000Z
DTSTAMP:20260404T131154Z
UID:LoReL/5
DESCRIPTION:Title: <a href="https://stable.researchseminars.org/talk/LoReL
 /5/">Gradualizing the Calculus of Inductive Constructions</a>\nby Éric Ta
 nter (Universidad de Chile\, Chile) as part of LoReL Seminar\n\n\nAbstract
 \n(joint work with M. Bertrand\, K. Maillard\, N. Tabareau)\n\nAcknowledgi
 ng the ordeal of a fully formal development in a proof assistant such as C
 oq\, we investigate gradual variations on the Calculus of Inductive Constr
 uction (CIC) for swifter prototyping with imprecise types and terms. We ob
 serve\, with a no-go theorem\, a crucial tradeoff between graduality and t
 he key properties of canonicity\, decidability and closure of universes un
 der dependent product that CIC enjoys. Beyond this Fire Triangle of Gradua
 lity\, we explore the gradualization of CIC with three different compromis
 es\, each relaxing one edge of the Fire Triangle. We develop a parametrize
 d presentation of Gradual CIC that encompasses all three variations\, and 
 jointly develop their metatheory. We first present a bidirectional elabora
 tion of Gradual CIC to a dependently-typed cast calculus\, which elucidate
 s the interrelation between typing\, conversion\, and graduality. We then 
 establish the metatheory of this cast calculus through both a syntactic mo
 del into CIC\, which provides weak canonicity\, confluence\, and when appl
 icable\, normalization\, and a monotone model that purports the study of t
 he graduality of two of the three variants. This work informs and paves th
 e way towards the development of malleable proof assistants and dependentl
 y-typed programming languages.\n\n[The talk will be on the “intro materi
 al” of the paper\, namely the analysis that leads to the Fire Triangle\,
  and not on the gory technical details of the metatheory of Gradual CIC]\n
LOCATION:https://stable.researchseminars.org/talk/LoReL/5/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Daniel Ventura (Universidade Federal de Goiás\, Brazil)
DTSTART:20210624T170000Z
DTEND:20210624T180000Z
DTSTAMP:20260404T131154Z
UID:LoReL/6
DESCRIPTION:Title: <a href="https://stable.researchseminars.org/talk/LoReL
 /6/">Skeletons Out! The Spirit of Node Replication</a>\nby Daniel Ventura 
 (Universidade Federal de Goiás\, Brazil) as part of LoReL Seminar\n\n\nAb
 stract\n(joint work with Delia Kesner and Loïc Peyrot)\n\nIn this talk we
  introduce a term calculus implementing higher-order node replication\, wh
 ere substitution of terms are executed constructor by constructor. Besides
  implementing the full node replication\, two evaluation strategies are co
 nsidered --call-by-name and fully lazy call-by-need-- based on the key not
 ion of skeleton\, with skeleton extraction internally codified in the calc
 ulus. Observational equivalence between strategies is proved through a sta
 ndard non-idempotent intersection type system.\n
LOCATION:https://stable.researchseminars.org/talk/LoReL/6/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Rafael Romero (Universidad de Buenos Aires\, Argentina)
DTSTART:20210701T170000Z
DTEND:20210701T180000Z
DTSTAMP:20260404T131154Z
UID:LoReL/7
DESCRIPTION:Title: <a href="https://stable.researchseminars.org/talk/LoReL
 /7/">A note on confluence in typed probabilistic lambda calculi</a>\nby Ra
 fael Romero (Universidad de Buenos Aires\, Argentina) as part of LoReL Sem
 inar\n\n\nAbstract\n(joint work with Alejandro Díaz-Caro)\n\nOn the topic
  of probabilistic rewriting\, there are several works studying both termin
 ation and confluence of different systems. While working with a lambda cal
 culus modelling quantum computation\, we found a system with probabilistic
  rewriting rules and strongly normalizing terms. We examine the effect of 
 small modifications in probabilistic rewriting\, affine variables\, and st
 rategies on the overall confluence in this strongly normalizing probabilis
 tic calculus.\n
LOCATION:https://stable.researchseminars.org/talk/LoReL/7/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Malena Ivnisky (Universidad de Buenos Aires\, Argentina)
DTSTART:20210708T170000Z
DTEND:20210708T180000Z
DTSTAMP:20260404T131154Z
UID:LoReL/8
DESCRIPTION:Title: <a href="https://stable.researchseminars.org/talk/LoReL
 /8/">A finite-dimensional model for affine\, linear quantum lambda calculi
  with general recursion</a>\nby Malena Ivnisky (Universidad de Buenos Aire
 s\, Argentina) as part of LoReL Seminar\n\n\nAbstract\n(joint work with Al
 ejandro Díaz-Caro\, Hernán Melgratti\, and Benoît Valiron)\n\nWe introd
 uce a concrete domain model for the quantum lambda calculus and lambda-rho
  extended with a fixpoint operator. A distinctive feature of lambda-rho is
  that it relies on density matrices for describing both quantum informatio
 n and probabilistic distributions over computation states. It has been sho
 wn that there is a conservative translation from lambda-rho to the quantum
  lambda calculus of Selinger and Valiron. In contrast to existing models f
 or quantum lambda calculi featuring recursion with intuitionistic arrows\,
  our model is finite-dimensional and does not need more than cones of posi
 tive matrices and affine arrows.\n
LOCATION:https://stable.researchseminars.org/talk/LoReL/8/
END:VEVENT
END:VCALENDAR
