Deposits of clastic carbonate-dominated (calciclastic) sedimentary slope systems in the rock record have been identified mostly as linearly-consistent carbonate apron deposits, even though most ancient clastic carbonate slope deposits fit the submarine fan systems better. Calciclastic submarine fans are consequently rarely described and are poorly understood. Subsequently, very little is known especially in mud-dominated calciclastic submarine fan systems. Presented in this study are a sedimentological core and petrographic characterisation of samples from eleven boreholes from the Lower Carboniferous of Bowland Basin (Northwest England) that reveals a >250 m thick calciturbidite complex deposited in a calciclastic submarine fan setting. Seven facies are recognised from core and thin section characterisation and are grouped into three carbonate turbidite sequences. They include: 1) Calciturbidites, comprising mostly of highto low-density, wavy-laminated bioclast-rich facies; 2) low-density densite mudstones which are characterised by planar laminated and unlaminated muddominated facies; and 3) Calcidebrites which are muddy or hyper-concentrated debrisflow deposits occurring as poorly-sorted, chaotic, mud-supported floatstones. These

Phd by thesis

The organization of behavior: A neuropsychological theory

Large language models have been shown to achieve remarkable performance across a variety of natural language tasks using few-shot learning , which drastically reduces the number of task-speciﬁc training examples needed to adapt the model to a particular application. To further our understanding of the impact of scale on few-shot learning, we trained a 540-billion parameter, densely activated, Transformer language model, which we call Pathways Language Model (PaLM). We trained PaLM on 6144 TPU v4 chips using Pathways, a new ML system which enables highly eﬃcient training across multiple TPU Pods. We demonstrate continued beneﬁts of scaling by achieving state-of-the-art few-shot learning results on hundreds of language understanding and generation benchmarks. On a number of these tasks, PaLM 540B achieves breakthrough performance, outperforming the ﬁnetuned state-of-the-art on a suite of multi-step reasoning tasks, and outperforming average human performance on the recently released BIG-bench benchmark. A signiﬁcant number of BIG-bench tasks showed discontinuous improvements from model scale, meaning that performance steeply increased as we scaled to our largest model. PaLM also has strong capabilities in multilingual tasks and source code generation, which we demonstrate on a wide array of benchmarks. We additionally provide a comprehensive analysis on bias and toxicity, and study the extent of training data memorization with respect to model scale. Finally, we discuss the ethical considerations related to large language models and discuss potential mitigation strategies.

PaLM: Scaling Language Modeling with Pathways

Minimal recursion semantics (MRS) is a framework for computational semantics that is suitable for parsing and generation and that can be implemented in typed feature structure formalisms. We discuss why, in general, a semantic representation with minimal structure is desirable and illustrate how a descriptively adequate representation with a nonrecursive structure may be achieved. MRS enables a simple formulation of the grammatical constraints on lexical and phrasal semantics, including the principles of semantic composition. We have integrated MRS with a broad-coverage HPSG grammar.

/pdf/minimal-recursion-semantics-an-introduction-4hdjxn0v8q.pdf

Minimal Recursion Semantics An Introduction

From Discourse to Logic

Although sloppy interpretation is usually accounted for by theories of ellipsis, it often arises in non-elliptical contexts. In this paper, a theory of sloppy interpretation is provided which captures this fact. The underlying idea is that sloppy interpretation results from a semantic constraint on parallel structures and the theory is shown to predict sloppy readings for deaccented and paycheck sentences as well as relational-, event-, and one-anaphora. It is further shown to capture the interaction of sloppy/strict ambiguity with quantification and binding. Finally, it is compared with other approaches to sloppy identity, in particular [4,12] [5]. 1 I n t r o d u c t i o n Sloppy interpretation involves two clauses: a source or antecedent clause and a t a r g e t clause that is, a clause containing a proform (the t a r g e t p r o f o r m ) . When the antecedent of the target proform also contains a proform (the source p r o f o r m ) , a sloppy interpretation may arise. In this case, the interpretation of the target proform differs from the interpretation of its source antecedent in the interpretation of the source proform. For instance, in: Example 1. Jon 1 [washes his: car] 2. Peter does2 too. the VP-ellipsis does has, among others, the sloppy interpretation washes Peter 's car (indeed, this is the preferred reading). The interpretation is sloppy because whereas in the source clause Jon washes his car the source proform his is interpreted as Jon, in the target clause Peter does too, it is re-interpreted as Peter. Although it is most often associated with VP-ellipsis, the phenomenon of sloppy identity is in fact very pervasive, and can occur in a wide range of configurations for instance: deaccenting (example 2 where the deaccented material is in bold face), paycheck sentences (example 3), VPE as a source proform (examples 4 and 5), non-pronominal referential elements involving implicit arguments (examples 6 and 7), event anaphora (examples 8 and 9) and one-anaphora (example i0):. ' The material occuring between brackets represents the interpretation being considered, in this case, the sloppy interpretation. Most examples are from [15].

Sloopy Identity

Stochastic dependency parsers can achieve very good results when they are trained on large corpora that have been manually annotated. Active learning is a procedure that aims at reducing this annotation cost by selecting as few sentences as possible that will produce the best possible parser. We propose a new selective sampling function for Active Learning that exploits two memory-based distances to find a good compromise between parser uncertainty and sentence representativeness. The reduced dependency between both parsing and selection models opens interesting perspectives for future models combination. The approach is validated on a French broadcast news corpus creation task dedicated to dependency parsing. It outperforms the baseline uncertainty entropy-based selective sampling on this task. We plan to extend this work with self- and co-training methods in order to enlarge this corpus and produce the first French broadcast news Tree Bank.

Memory-Based Active Learning for French Broadcast News

In the medical field, we have seen the emergence of health-bots that interact with patients to gather data and track their state. One of the downstream application is automatic questionnaire filling, where the content of the dialog is used to automatically fill a pre-defined medical questionnaire. Previous work has shown that answering questions from the dialog context can successfully be cast as a Natural Language Inference (NLI) task and therefore benefit from current pre-trained NLI models. However, NLI models have mostly been trained on text rather than dialogs, which may have an influence on their performance. In this paper, we study the influence of content transformation and content selection on the questionnaire filling task. Our results demonstrate that dialog pre-processing can significantly improve the performance of zero-shot questionnaire filling models which take health-bots dialogs as input.

Exploring the Influence of Dialog Input Format for Unsupervised Clinical Questionnaire Filling

Ce travail prend place dans le cadre plus general du developpement d'une plate-forme d'analyse syntaxique du francais parle. Nous decrivons la conception d'un modele automatique pour resoudre le lien anaphorique present dans les dislocations a gauche dans un corpus de francais parle radiophonique. La detection de ces structures devrait permettre a terme d'ameliorer notre analyseur syntaxique en enrichissant les informations prises en compte dans nos modeles automatiques. La resolution du lien anaphorique est realisee en deux etapes : un premier niveau a base de regles filtre les configurations candidates, et un second niveau s'appuie sur un modele appris selon le critere du maximum d'entropie. Une evaluation experimentale realisee par validation croisee sur un corpus annote manuellement donne une F-mesure de l'ordre de 40%.

https://hal.archives-ouvertes.fr/hal-00600510/document

Vers la détection des dislocations à gauche dans les transcriptions automatiques du Français parlé (Towards automatic recognition of left dislocation in transcriptions of Spoken French)

Dans cet article, nous presentons une architecture logicielle libre et ouverte pour le developpement de grammaires d'arbres adjoints a portee semantique. Cette architecture utilise un compilateur de metagrammaires afin de faciliter l'extension et la maintenance de la grammaire, et integre un module de construction semantique permettant de verifier la couverture aussi bien syntaxique que semantique de la grammaire. Ce module utilise un analyseur syntaxique tabulaire genere automatiquement a partir de la grammaire par le systeme DyALog. Nous presentons egalement les resultats de l'evaluation d'une grammaire du francais developpee au moyen de cette architecture.

https://hal.inria.fr/inria-00160393/document

Claire Gardent

Papers

Sloopy Identity

Memory-Based Active Learning for French Broadcast News

Exploring the Influence of Dialog Input Format for Unsupervised Clinical Questionnaire Filling

Vers la détection des dislocations à gauche dans les transcriptions automatiques du Français parlé (Towards automatic recognition of left dislocation in transcriptions of Spoken French)

SemTAG, une architecture pour le développement et l’utilisation de grammaires d’arbres adjoints à portée sémantique