A program denotes computations in some universe of objects. Abstract interpretation of programs consists in using that denotation to describe computations in another universe of abstract objects, so that the results of abstract execution give some information on the actual computations. An intuitive example (which we borrow from Sintzoff [72]) is the rule of signs. The text -1515 * 17 may be understood to denote computations on the abstract universe {(+), (-), (±)} where the semantics of arithmetic operators is defined by the rule of signs. The abstract execution -1515 * 17 → -(+) * (+) → (-) * (+) → (-), proves that -1515 * 17 is a negative number. Abstract interpretation is concerned by a particular underlying structure of the usual universe of computations (the sign, in our example). It gives a summary of some facets of the actual executions of a program. In general this summary is simple to obtain but inaccurate (e.g. -1515 + 17 → -(+) + (+) → (-) + (+) → (±)). Despite its fundamentally incomplete results abstract interpretation allows the programmer or the compiler to answer questions which do not need full knowledge of program executions or which tolerate an imprecise answer, (e.g. partial correctness proofs of programs ignoring the termination problems, type checking, program optimizations which are not carried in the absence of certainty about their feasibility, …).

/pdf/abstract-interpretation-a-unified-lattice-model-for-static-2jfaix0dre.pdf

Abstract interpretation: a unified lattice model for static analysis of programs by construction or approximation of fixpoints

For single databases, primary hindrances for end-user access are the volume of data that is becoming available, the lack of abstraction, and the need to understand the representation of the data. When information is combined from multiple databases, the major concern is the mismatch encountered in information representation and structure. Intelligent and active use of information requires a class of software modules that mediate between the workstation applications and the databases. It is shown that mediation simplifies, abstracts, reduces, merges, and explains data. A mediator is a software module that exploits encoded knowledge about certain sets or subsets of data to create information for a higher layer of applications. A model of information processing and information system components is described. The mediator architecture, including mediator interfaces, sharing of mediator modules, distribution of mediators, and triggers for knowledge maintenance, are discussed. >

Mediators in the architecture of future information systems

Author(s): Asanovic, K; Bodik, R; Catanzaro, B; Gebis, J; Husbands, P; Keutzer, K; Patterson, D; Plishker, W; Shalf, J; Williams, SW | Abstract: The recent switch to parallel microprocessors is a milestone in the history of computing. Industry has laid out a roadmap for multicore designs that preserves the programming paradigm of the past via binary compatibility and cache coherence. Conventional wisdom is now to double the number of cores on a chip with each silicon generation. A multidisciplinary group of Berkeley researchers met nearly two years to discuss this change. Our view is that this evolutionary approach to parallel hardware and software may work from 2 or 8 processor systems, but is likely to face diminishing returns as 16 and 32 processor systems are realized, just as returns fell with greater instruction-level parallelism. We believe that much can be learned by examining the success of parallelism at the extremes of the computing spectrum, namely embedded computing and high performance computing. This led us to frame the parallel landscape with seven questions, and to recommend the following: • The overarching goal should be to make it easy to write programs that execute efficiently on highly parallel computing systems • The target should be 1000s of cores per chip, as these chips are built from processing elements that are the most efficient in MIPS (Million Instructions per Second) per watt, MIPS per area of silicon, and MIPS per development dollar. • Instead of traditional benchmarks, use 13 “Dwarfs” to design and evaluate parallel programming models and architectures. (A dwarf is an algorithmic method that captures a pattern of computation and communication.) • “Autotuners” should play a larger role than conventional compilers in translating parallel programs. • To maximize programmer productivity, future programming models must be more human-centric than the conventional focus on hardware or applications. • To be successful, programming models should be independent of the number of processors. • To maximize application efficiency, programming models should support a wide range of data types and successful models of parallelism: task-level parallelism, word-level parallelism, and bit-level parallelism. 1 The Landscape of Parallel Computing Research: A View From Berkeley • Architects should not include features that significantly affect performance or energy if programmers cannot accurately measure their impact via performance counters and energy counters. • Traditional operating systems will be deconstructed and operating system functionality will be orchestrated using libraries and virtual machines. • To explore the design space rapidly, use system emulators based on Field Programmable Gate Arrays (FPGAs) that are highly scalable and low cost. Since real world applications are naturally parallel and hardware is naturally parallel, what we need is a programming model, system software, and a supporting architecture that are naturally parallel. Researchers have the rare opportunity to re-invent these cornerstones of computing, provided they simplify the efficient programming of highly parallel systems.

/pdf/the-landscape-of-parallel-computing-research-a-view-from-1i5pswggnf.pdf

The Landscape of Parallel Computing Research: A View from Berkeley

There has been increased research interest in systems composed of multiple autonomous mobile robots exhibiting cooperative behavior. Groups of mobile robots are constructed, with an aim to studying such issues as group architecture, resource conflict, origin of cooperation, learning, and geometric problems. As yet, few applications of cooperative robotics have been reported, and supporting theory is still in its formative stages. In this paper, we give a critical survey of existing works and discuss open problems in this field, emphasizing the various theoretical issues that arise in the study of cooperative robotics. We describe the intellectual heritages that have guided early research, as well as possible additions to the set of existing motivations.

Cooperative mobile robotics: antecedents and directions

Federated Learning is a distributed machine learning approach which enables model training on a large corpus of decentralized data. We have built a scalable production system for Federated Learning in the domain of mobile devices, based on TensorFlow. In this paper, we describe the resulting high-level design, sketch some of the challenges and their solutions, and touch upon the open problems and future directions.

Towards Federated Learning at Scale: System Design

This paper proposes a modular ACTOR architecture and definitional method for artificial intelligence that is conceptually based on a single kind of object: actors [or, if you will, virtual processors, activation frames, or streams]. The formalism makes no presuppositions about the representation of primitive data structures and control structures. Such structures can be programmed, micro-coded, or hard wired in a uniform modular fashion. In fact it is impossible to determine whether a given object is "really" represented as a list, a vector, a hash table, a function, or a process. The architecture will efficiently run the coming generation of PLANNER-like artificial intelligence languages including those requiring a high degree of parallelism. The efficiency is gained without loss of programming generality because it only makes certain actors more efficient; it does not change their behavioral characteristics. The architecture is general with respect to control structure and does not have or need goto, interrupt, or semaphore primitives. The formalism achieves the goals that the disallowed constructs are intended to achieve by other more structured methods. PLANNER Progress "Programs should not only work, but they should appear to work as well." PDP-1X Dogma The PLANNER project is continuing research in natural and effective means for embedding knowledge in procedures. In the course of this work we have succeeded in unifying the formalism around one_ fundamental concept: the ACTOR. Intuitively, an ACTOR is an active agent which plays a role on cue according to a script. We use the ACTOR metaphor to emphasize the inseparability of control and data flow in our model. Data structures, functions, semaphores, monitors, ports, descriptions, Quillian nets, logical formulae, numbers, identifiers, demons, processes, contexts, and data bases can all be shown to be special cases of actors. All of the above are objects with certain useful modes of behavior. Our formalism shows how all of the modes of behavior can be defined in terms of one kind of behavior: sending messages to actors. An actor is always invoked uniformly in exactly the same way regardless of whether it behaves as a recursive function, data structure, or process. "It is vain to multiply Entities beyond need." William of Occam "Monotheism is the Answer." The unification and simplification of the formalisms for the procedural embedding of knowledge has a great many benefits for us: FOUNDATIONS: The concept puts procedural semantics [the theory of how things operate] on a firmer basis. It will now be possible to do cleaner theoretical studies of the relation between procedural semantics and set-theoretic semantics such as model theories of the quantificational calculus and the lambda calculus. LOGICAL CALCULAE: A procedural semantics is developed for the quantificational calculus. The logical constants FOR-ALL, THERE-EXISTS, AND, OR, NOT, and IMPLIES are defined as actors. KNOWLEDGE BASED PROGRAMMING is programming in an environment which has a substantial knowledge base in the application area for which the programs are intended. The actor formalism aids knowledge based programming in the following ways: PROCEDURAL EMBEDDING of KNOWLEDGE, TRACING BEHAVIORAL DEPENDENCIES, and SUBSTANTIATING that ACTORS SATISFY their INTENTIONS. INTENTIONS: Furthermore the confirmation of properties of procedures is made easier and more uniform. Every actor has an INTENTION which checks that the prerequisites and the context of the actor being sent the message are satisfied. The intention is the CONTRACT that the actor has with the outside world. How an actor fullfills its contract is its own business. By a SIMPLE BUG we mean an actor which does not satisfy its intention. We would like to eliminate simple debugging of actors by the META-EVALUATION of actors to show that they satisfy their intentions. Suppose that there is an external audience of actors E which satisfy the intentions.of the actors to which they send messages. Intuitively, the principle of ACTOR INDUCTION states that the intentions of all actions caused by E are in turn satisfied provided that the following condition holds: If for each actor A the' intention of A is satisfied => that the intentions of all actors sent messages by A are satisfied. Computational induction [Manna], structural induction [Burstall], and Peano induction are all special cases of ACTOR induction. Actor based intentions have the following advantages: The intention is decoupled from the actors it describes. Intentions of concurrent actions are more easily disentangled. We can more elegantly write intentions The intentions are written in the same formalism as the Because for dialogues between actors. procedures they describe. Thus for example intentions can have intentions, protection is an intrinsic property of actors, we hope to be able to deal with protection issues in the same straight forward manner as more conventional intentions. Intentions of data structures are handled by the same machinery as for all other actors. COMPARATIVE SCHEMATOLOGY: The theory of comparative power of control structures is I

/pdf/a-universal-modular-actor-formalism-for-artificial-3nnj28hg51.pdf

A universal modular ACTOR formalism for artificial intelligence

The PLANNER project is continuing research in natural and effective means for embedding knowledge in procedures. In the course of this work we have succeeded in unifying the formalism around one fundamental concept: the ACTOR. Intuitively, an ACTOR is an active agent which plays a role on cue according to a script. We use the ACTOR metaphor to emphasize the inseparability of control and data flow in our model. Data structures, functions, semaphores, monitors, ports, descriptions, Quillian nets, logical formulae, numbers, identifiers, demons, processes, contexts, and data bases can all be shown to be special cases of actors. All of the above are objects with certain useful modes of behavior. Our formalism shows how all of these modes of behavior can be defined in terms of one kind of behavior: sending messages to actors. An actor is always invoked uniformly in exactly the same way regardless of whether it behaves as a recursive function, data structure, or process.

Actor induction and meta-evaluation

Knowledge Based Programming is programming in an environment which has substantial knowledge of the semantic domain for which the programs are being written and of the purposes that the programs are supposed to satisfy. Actors are a semantic concept in which no active process is ever allowed to treat anything as an object; instead a polite request must be extended to accomplish what the activator desires. Using actors the PLANNER Project is constructing a Programming Apprentice to make it easier for expert programmers to do knowledge based programming. The apprentice is to aid in establishing and maintaining consistency of specifications, validating that modules meet their specifications, answering questions about behavioral dependencies between modules, and analyzing the implications of perturbations in modules and their specifications.

Richard Steiger

Papers

A universal modular ACTOR formalism for artificial intelligence

Actor induction and meta-evaluation

Behavioral semantics of nonrecursive control structures