No abstract

The question of how to probe contextual word representations for linguistic structure in a way that is both principled and useful has seen significant attention recently in the NLP literature. In our contribution to this discussion, we argue for a probe metric that reflects the fundamental trade-off between probe complexity and performance: the Pareto hypervolume. To measure complexity, we present a number of parametric and non-parametric metrics. Our experiments using Pareto hypervolume as an evaluation metric show that probes often do not conform to our expectations-e.g., why should the non-contextual fastText representations encode more morpho-syntactic information than the contextual BERT representations? These results suggest that common, simplistic probing tasks, such as part-of-speech labeling and dependency arc labeling, are inadequate to evaluate the linguistic structure encoded in contextual word representations. This leads us to propose full dependency parsing as a probing task. In support of our suggestion that harder probing tasks are necessary, our experiments with dependency parsing reveal a wide gap in syntactic knowledge between contextual and non-contextual representations.

Measuring what linguistic information is encoded in neural models of language has become popular in NLP. Researchers approach this enterprise by training "probes"supervised models designed to extract linguistic structure from another model's output. One such probe is the structural probe (Hewitt and Manning, 2019), designed to quantify the extent to which syntactic information is encoded in contextualised word representations. The structural probe has a novel design, unattested in the parsing literature, the precise benefit of which is not immediately obvious. To explore whether syntactic probes would do better to make use of existing techniques, we compare the structural probe to a more traditional parser with an identical lightweight parameterisation. The parser outperforms structural probe on UUAS in seven of nine analysed languages, often by a substantial amount (e.g. by 11.1 points in English). Under a second less common metric, however, there is the opposite trend-the structural probe outperforms the parser. This begs the question: which metric should we prefer?

A longstanding debate in semiotics centers on the relationship between linguistic signs and their corresponding semantics: is there an arbitrary relationship between a word form and its meaning, or does some systematic phenomenon pervade? For instance, does the character bigram gl have any systematic relationship to the meaning of words like glisten, gleam and glow? In this work, we offer a holistic quantification of the systematicity of the sign using mutual information and recurrent neural networks. We employ these in a data-driven and massively multilingual approach to the question, examining 106 languages. We find a statistically significant reduction in entropy when modeling a word form conditioned on its semantic representation. Encouragingly, we also recover wellattested English examples of systematic affixes. We conclude with the meta-point: Our approximate effect size (measured in bits) is quite small-despite some amount of systematicity between form and meaning, an arbitrary relationship and its resulting benefits dominate human language.

We present methods for calculating a measure of phonotactic complexity-bits per phoneme-that permits a straightforward cross-linguistic comparison. When given a word, represented as a sequence of phonemic segments such as symbols in the international phonetic alphabet, and a statistical model trained on a sample of word types from the language, we can approximately measure bits per phoneme using the negative log-probability of that word under the model. This simple measure allows us to compare the entropy across languages, giving insight into how complex a language's phonotactics are. Using a collection of 1016 basic concept words across 106 languages, we demonstrate a very strong negative correlation of −0.74 between bits per phoneme and the average length of words.

The uniform information density (UID) hypothesis posits a preference among language users for utterances structured such that information is distributed uniformly across a signal. While its implications on language production have been well explored, the hypothesis potentially makes predictions about language comprehension and linguistic acceptability as well. Further, it is unclear how uniformity in a linguistic signal-or lack thereof-should be measured, and over which linguistic unit, e.g., the sentence or language level, this uniformity should hold. Here we investigate these facets of the UID hypothesis using reading time and acceptability data. While our reading time results are generally consistent with previous work, they are also consistent with a weakly super-linear effect of surprisal, which would be compatible with UID's predictions. For acceptability judgments, we find clearer evidence that non-uniformity in information density is predictive of lower acceptability. We then explore multiple operationalizations of UID, motivated by different interpretations of the original hypothesis, and analyze the scope over which the pressure towards uniformity is exerted. The explanatory power of a subset of the proposed operationalizations suggests that the strongest trend may be a regression towards a mean surprisal across the language, rather than the phrase, sentence, or document-a finding that supports a typical interpretation of UID, namely that it is the byproduct of language users maximizing the use of a (hypothetical) communication channel. 1

Abstract-d Caves on other planetary bodies offer sheltered habitat for future human explorers and numerous clues to a planet's past for scientists. While recent orbital imagery provides exciting new details about cave entrances on the Moon and Mars, the interiors of these caves are still unknown and not observable from orbit. Multi-robot teams offer unique solutions for exploration and modeling subsurface voids during precursor missions. Robot teams that are diverse in terms of size, mobility, sensing, and capability can provide great advantages, but this diversity, coupled with inherently distinct low-level behavior architectures, makes coordination a challenge. This paper presents a framework that consists of an autonomous frontier and capability-based task generator, a distributed market-based strategy for coordinating and allocating tasks to the different team members, and a communication paradigm for seamless interaction between the different robots in the system. Robots have different sensors, (in the representative robot team used for testing: 2D mapping sensors, 3D modeling sensors, or no exteroceptive sensors), and varying levels of mobility. Tasks are generated to explore, model, and take science samples. Based on an individual robot's capability and associated cost for executing a generated task, a robot is autonomously selected for task execution. The robots create coarse online maps and store collected data for high resolution offline modeling. The coordination approach has been field tested at a mock cave site with highly-unstructured natural terrain, as well as an outdoor patio area. Initial results are promising for applicability of the proposed multi-robot framework to exploration and modeling of planetary caves.

2 R R R RESUMO ESUMO ESUMO ESUMOO objetivo do estudo é evidenciar e discutir a configuração dos discursos sobre a responsabilidade social nas organizações, assim como a sua incorporação da temática ambiental. Assume-se que várias estratégias são adotadas para disseminar determinado discurso sobre essas questões. Porém as ambigüidades inerentes à inserção desses temas nas organizações remetem a um discurso fragmentado (Fineman, 1996), revelando práticas de abertura e de dissimulação. A discussão teórica parte do tema responsabilidade social, o confronta com a temática ambiental e relaciona os discursos com as ambigüidades das práticas organizacionais oriundas dos dois temas. Por fim, um estudo de caso na Antena embasa a discussão. Os dados foram coletados por meio de pesquisa documental e de 40 entrevistas semi-estruturadas. Foi utilizado o método da Análise do Discurso (Fiorin, 1989). Como conclusão, comprovou-se que a preocupação com a responsabilidade social, já incorporando a temática ambiental, permeia a organização. Ela está na fala e em ações da alta direção, dos gerentes e de boa parte dos técnicos. Entretanto o silêncio sobre os limites dessa responsabilidade é preenchido por um grupo de técnicos que revela a dissimulação, quando a abertura ameaça seus objetivos específicos.Palavras-chave: responsabilidade social; discursos; temática ambiental.