We describe new algorithms for training tagging models, as an alternative to maximum-entropy models or conditional random fields (CRFs). The algorithms rely on Viterbi decoding of training examples, combined with simple additive updates. We describe theory justifying the algorithms through a

. If the validation succeeds, the code is guaranteed to respect the safety policy without relying on run-time checks. The main practical difficulty of PCC is in generating the safety proofs. In order to gain some preliminary experience with this, we have written several network packet filters in hand-tuned DEC Alpha

between a template matching and a maximum pooling operation. We demonstrate the strength of the approach on a range of recognition tasks: From invariant single object recognition in clutter to multiclass categorization problems and complex scene understanding tasks that rely on the recognition of both

present two solutions with similar structure. The first one is privately verifiable and builds elegantly on pseudorandom functions (PRFs); the second allows for publicly verifiable proofs and is built from the signature scheme of Boneh, Lynn, and Shacham in bilinear groups. Both solutions rely

guarantees to provide in return. This paper presents a proof technique that permits us to infer that a program P satisfies a rely/guarantee specification R oe G, given that we know P satisfies a finite collection of rely/guarantee specifications R i oe G i ; (i 2 I). The utility of the proof technique

allows fine-grained interleaving and nested concurrency; it is defined by an operational semantics; the proof that the rely/guarantee rules are consistent with that semantics (including termination) is by a structural induction. A key lemma which relates the states which can arise from the extra

dimensions under weak requirements on the switching and failing communication topology. The correctness proof relies on proximity graphs and their properties and on a LaSalle Invariance Principle for nondeterministic discrete-time systems.

hiding to concurrency verification. Our logic, for the first time, supports a frame rule over rely/guarantee conditions so that specifications of program modules only need to talk about the resources used locally, and the verified modules can be reused in different threads without redoing the proof

We prove Sen’s third conjecture that there are no on-shell perturbative excitations of the tachyon vacuum in open bosonic string field theory. The proof relies on the existence of a special state A, which, when acted on by the BRST operator at the tachyon vacuum, gives the identity. While

Abstract. The Rely-Guarantee approach is a well-known compositional method for proving Hoare logic properties of concurrent programs. In this approach, predicates in the proof must be proved invariant (or stable) under interference from the environment. We describe a framework, and a prototype