Over the last decade, Geosciences have undergone a major technological revolution. Gone are the days when Earth Science was limited to hammering rocks in the field and producing maps with old-school surveying tools. The emergence of high-performance computational technologies, Big Data driven geo-analytics and 3D visualisations are now at the heart of this field. New understanding of the evolution and interrelationships between the Earth subsystems (Earth's interior-ocean-atmosphere) and human activities makes it imperative that we, as teachers, provide high quality science education to all students in order to form the next generation of scientifically literate citizen. I believe Geosciences have a central role to play in this educational mission, particularly with regard to natural hazards, resource use, anthropogenic impacts and environmental awareness.This reflective portfolio puts forward my teaching & learning vision to engage undergraduate students in genuine inquiry and provide them with the opportunity to experience the process of doing science. To do so, I believe, we must design our teaching materials based on research in human learning and improve awareness and engagement in the classroom with emerging technologies. I think this integrated approach should provide the knowledge base, methodologies, and global context that can make science accessible, relevant and meaningful for all students.\(\)Teaching & learning visionOur personal experiences influence our way of learning & teaching. These processes have nothing to do with a supposed innate nature. Knowledge acquisition and understanding are complex and remain a controversial topic among educators and education researchers. However, our understanding of how students learn has greatly improved in the past two decades \cite{Manduca_2016}. It is now established that students are most effective in acquiring science knowledge when they use a range of cognitive processes, including posing questions, using knowledge and scientific principles for solving problems, and conveying understanding of complex issues to others. In my opinion, the teacher's role is to nurture these skills, control the learning process as well as evaluate the learners' work through adapted tools. Effective learning goes beyond the memorization of facts. Accordingly Geosciences-based education should be designed to effectively educate students by (1) building genuine inquiry and the excitement of discovery into my courses, and (2) giving students experience in the process of science \cite{Murray_2015,LaDue_2015,Kortz_2008}.Many Geosciences courses to date do not fulfil these criteria and are instead characterised by the prevalence of passive rather than active learning, an emphasis on factual knowledge with no experience in the process of science, "cookbook" rather than inquiry-based labs, and a lack of relevance of course material \cite{LaDue_2015,Kortz_2008,Kenyon_2000}. I believe that for effective learning, educators need to go beyond coverage of a certain body of content and incorporate genuine inquiry and hands on experience in the classroom as well as teaching communication skills, teamwork, critical thinking, and lifelong learning skills.To promote this broader perspective on student learning I think we need to (i) decrease emphasis on fact-focused, lecture-style courses, (ii) increase emphasis on actively engaging students during class time, (iii) support better integration of research and research-like experiences, (iv) emphase on improving oral and written communication skills, and working in teams to solve problems.Courses in Geosciences are also uniquely suited to drawing connections with societal issues and making science relevant to all students. Field studies are vital in connecting the science concepts to students life experiences and should involve students in questioning, data gathering, and interpretation \cite{McConnell_2005,Reynolds_1998}. Computerised virtual field trips have proven to be efficient within the School of Geosciences to help students understand a field area in a larger context and to focus students' attention on particularly important observations (such as the VirtualGeoLife, the Magic Planet as well as the Mines, Wines & Thoroughbreds virtual fieldtrips). I strongly believe that our courses should take advantage of the revolution in multimedia and information technology. Computers allow students access to large spatial and time series databases. This requires user-friendly display and open-source analysis tools for students. Modelling and simulation software provide unique opportunities for students to study systems thinking and to compare models output with real data.Assessment is also critical to demonstrate that student learning has been improved and course goals have been achieved. The goals, strategies, and outcomes of my teaching approach have to be critically evaluated, with rigorous assessment of classroom materials, pedagogy, and student learning. Student performance needs to recognise and accommodate academic, cultural, and learning-style diversity especially here at the University of Sydney.  We also need to convey the fact that mistakes are positive and participation in the classroom should be encouraged \cite{Kenyon_2000}. They provide a better way to learn since a long term commitment from the student is required to become a professional geoscientist. For assessment to be efficient, I also believe that individual courses as well as the overall curriculum need to contain well-defined educational goals for student learning. Also, I think that teachers should maintain an ongoing dialogue with students to evaluate the effectiveness of their teaching practices.Integrate research & educationI am a strong advocate for the integration of research and education. The willingness to explore the unknown and the excitement about challenging and open-ended investigations stem from research experiences. Several studies have shown that integrating research and education is a powerful technique which provide students with the skills in problem solving and critical thinking that will be necessary in their future professional life \cite{Kanim_2001,nsf00,Singer_2013}.Research provides the opportunity for students not just to learn about science but to do science. Without such experiences, students gain knowledge but little practical experience in applying concepts and methodology that is fundamental to scientific work. Research experiences provide students with a greater understanding of scientific investigation, increases their scientific literacy, and promotes free thinking and creative approaches to problem solving \cite{2006}. Students who are able to put scientific concepts and facts into practice through research have a better understanding of the processes and the dynamic nature of scientific knowledge.Student research needs to be defined as inquiry-based study, discovery, or rediscovery resulting in original contribution \cite{Hall_2013,2011}. In my views, student research needs to be articulated around (i) hypothesis formulation, (ii) collection and use of real-time data and other research materials to test hypotheses, and (iii) analysis of data in individual or group settings. I think that inquiry-based exercises can be developed at all levels of the curriculum, from the introductory course to a senior capstone research project. Within the School of Geosciences, research includes both Geology, Geophysics & Geography fields, including projects investigating physical and human/societal impact. To be efficient, integrated research in learning curriculum needs to include the analysis of the difficulties that students have in thinking about and reasoning with Earth science concepts to approach a problem or conduct scientific research.Self-responsibility and a sense of ownership of a project are natural outcomes of the research experience; this enhances students' competence and confidence and allows for additional assessment of their skills and abilities \cite{Kolb_2005}. It also keeps curriculum current, improves pedagogical outcome and increases academic's connection with students. I also believe that this integration is helpful to the recruitment and retention of undergraduates. Furthermore, the research process could also promote interdisciplinary interactions between research teams or different schools. Therefore I think it is critical that educators design discovery-based exercises across the curriculum to allow students to develop these important skills and, in turn, to potentially stimulate new research directions.Putting my  vision into practiceInterdisciplinary unitsAs mentioned above, Geosciences may be the most interdisciplinary of all STEM disciplines. Earth system sciences and the complex sub-systems of the cryosphere, the atmosphere, the lithosphere, the biosphere, and the hydrosphere subsume all human activity and are critical to every aspects of life on Earth. Over the last two years, I have been working with my colleagues to propose in two of our units of study (MARS5001 & GEOS3009) an interdisciplinary curriculum focussing on the integration of marine dynamics, coral reef evolution, sustainable resource and coastal management. By promoting interdisciplinary thinking, students are exploring questions that are not confined to a single discipline and they recognise the need to solve problems which solutions are beyond the scope of a single area of research practice. This requires that students are able to seamlessly integrate multiple perspectives  by integrating information, data techniques, tools, concepts and/or theories from several bodies of specialised knowledge \cite{Kezar_2012}. Despite the challenging nature of such courses, students, especially undergraduates, are strongly attracted to these courses, mainly because of their societal relevance.    Learning in the field has long held a prominent role in the education of geoscientists \cite{Bishop_2009}. I strongly believe field experiences are crucial to both learning and professional preparation. Fieldwork needs to be valued mainly for perceived cognitive gains, such as knowledge and understanding, and for enabling learners to interact with geological phenomena in their natural state \cite{Petcovic_2014}.  This year in the Coastal Environments & Processes unit of study (GEOS3009), I organised with my colleagues a 3-days coastal field trip in the Kioloa Campus. From students feedbacks it is clear that they recognised the value of field experiences in enhancing their geologic knowledge and problem-solving skills. It allows them to make their own observations, order their experiences, make decisions and set their own priorities as to what to focus on and what to ignore.  I believe that it helps them in becoming autonomous and self-directed learners. I also think that field-based learning is enhanced when students use their own field dataset to process, plot, model and interpret their findings throughout the semester when additional quantitative techniques (from laboratory experiments and numerical tools) are introduced. It allows them to develop critical thinking when assessing alternative hypotheses. For this to be effective it requires that our students not only gain field-based knowledge but also possess the ability to interpret their results with numerical tools. These skills are common to most STEM students and need to be nurtured throughout the undergraduates programs. With colleagues from SOLES, Physics and Mathematics I have been developing an Open Learning Environment  module (2016 Strategic Education Grant #16054  Analysing and plotting data with R/Python) to foster computational literacy in classroom and ensure that our students have the necessary skills to succeed in their studies but also in their professional life.Learning supportOver the last 2 years, I followed the Peer Observation and Review of Teaching program as a volunteer, first, and then during the Graduate Certificate in Educational Studies (Higher Education). As an observer I gained several techniques and approaches to learning that have significantly changed and improved my own teaching. From this reflective experience, I have redesigned most of my teaching materials embedding more interactive contents to improve students understanding of important concepts, and to offer an intuitive approach to complex problems \cite{Sitzmann}. My lectures and practicals are available online on my personal teaching website GeosLearn. I think digital learning resources support information processing by helping students to develop mental representations through the mix of media elements presented to them.  Most of the units of study I teach embed classical lectures and online information which combine multimedia elements including text, image, and video. The main purposes of these resources are to introduce students to some of the topics discussed in the classroom and to allow them to work at their own pace as an extension activity. These materials are often associated to practicals and hands-on exercises that meet the content and learning objectives curricula. To enhance the development of independent and autonomous learning, I sometimes provide specific exercises that go beyond the course objectives but contribute to course-specific learning outcomes.Transformative teachingMany of the scientific software packages routinely used in our University are proprietary and closed-source, preventing the students from having a complete understanding of the final scientific results. By promoting the use of open-source based technologies, we could change the educational process for new scientists, providing problem contextualization within notebooks and extending the interactive experience beyond a passive interrogation mode to a more active learning approach. To achieve this vision for Geosciences, I proposed back in 2015 to deploy a new, open-source and collaborative technological infrastructure allowing students to learn through the interactive and collaborative exploration of data and models, and incrementally build computational and data analysis expertise critical to form the next generation of scientifically literate citizen. This infrastructure, called Jupyter \cite{perez15} and whose early adopters include Berkeley and MIT \cite{Hamrick_2016}, has the potential to serve all students, to be truly transformative, and to give our graduates a tangible competitive advantage in their professional life.Jupyter has opened-up a radically new way to deliver access to numerical and data processing tools to the community. It allows access to over 50 programming languages (including Python, Julia, R and many others). Once installed on a server, students  can open teaching materials called  notebooks embedding explanatory texts, executable codes and graphic outputs. These notebooks can be shared with the community to form the basis for collaborative work. With immediate feedback, flexible visualisation, easy access to docstrings and ability to explore modules Jupyter environment lends itself to teaching and learning. Today, it serves not only the academic and scientific communities, but also a much broader constituency of data scientists in research, industry and journalism \cite{Shen_2014}.University-wide, several academics from other Schools (Mathematics, IT, Agriculture to cite a few) were using Jupyter environment for research purposes only. During the second semester of 2015, with the help of the ICT TechLab \cite{salles15a}, I was the first in our University to design a Jupyter-based teaching course and ran a successful pilot within the School of Geosciences in three units of study. Now, many of the core senior units I am teaching in Geology & Marine Sciences are articulated around this new technology. I generally use Jupyter in both lectures and practicals but for two distinct purposes:During lectures, Jupyter notebooks are learning supports making the learning more interactive. For example rather than having a static diagram or graph in a slide, I will include codes to generate 2D plotting or 3D visualisation so that students can discuss and understand how various parameters will change a specific process. I believe it enables students to relate theoretical ideas to everyday practise and fosters an emphasis to deep-learning. Another key use in lecture relates to complex problems solving. Standard approaches to teaching physical oceanography or geophysics rely on analytical (pen-and-paper) solution of problems, but are limited to only a few examples which can be solved by undergraduates. Jupyter allows complex, real-life problems to be tackled, again giving new, deeper insight into the subject being studied, and making the learning process active. In addition, it also empowers students with tools that can perform calculations that parallel current research.During practicals, the students are involved in active inquiry through interactive and reflective Jupyter notebooks, interpretation and independent research. It follows Kolb’s four stages of the learning cycle: experience, reflection, conceptualisation and experimentation \cite{Kolb_2005}. It also emphasises the importance of usability, collaboration and reproducibility.Students participation in classroom, students interviews as well as feedbacks from UoS surveys demonstrate that this technology enhances active and participatory learning and promotes engaged enquiry. Retrospectively, I have noticed that active student engagement through Jupyter environment has triggered positive emotional and learning responses leading to improved motivation and effectiveness in achieving learning outcomes. This aligns closely with the Science Threshold Learning Outcomes \cite{kirkup13} statements on inquiry and problem solving. Jupyter offers inclusive and open learning environments for the extension of knowledge and skills by providing flexible options for investigating problems in a vast range of disciplines and encouraging peer interaction. This year with my colleagues from the School of IT, Physics,  Mathematics & Agriculture we are submitting a grant to offer Jupyter as an inter-Faculty learning tool for STEM students. Wrap upMy vision for University graduates is to develop an in depth disciplinary expertise, served by broad skills in critical thinking and problem solving, good communication skills, and a strong expertise in the use of information and digital technologies. This reflective portfolio emphasises the need to link teaching and learning to the real world, to provide opportunities for interdisciplinary studies, and to better align our teaching with our research. To achieve this vision I have worked on several strategies based on (1) multiple learning resources (tutorials, notebooks) freely accessible to enhance students experience through easier access to eLearning spaces and foster participation, (2) the use of web-based programming environment (namely Jupyter) which significantly increases the accessibility of my materials and promotes an uniform and integrated approach to computing university-wide, and finally (3) the development of interdisciplinary units of study to help students connect and integrate knowledge and skills from across disciplines to solve complex problems and prepare them for future learning as lifelong learners in their careers and as citizens.\(\)

Introduction to Goal-directed BehaviourBehaviour in humans is intrinsically goal-oriented. In all realistic behaviour, goals are achieved by assembling a series of sub-tasks, each separately defined and solved (Duncan, 2010). To effectively guide behaviour in pursuit of goals, mental representations of appropriate lines of action must be maintained and executed. This occurs in all forms of behaviour, whether the situation is familiar or whether it is novel. In familiar situations, the rules governing the environmental context are over-prepared, so little must be controlled due to the complexity, danger, or novelty of the situation. These situations typically bring forth multiple or competing routes of potential action which have not been explored cognitive effort is required. Yet rule-based errors can still occur, especially when there is a slight change in the environment. Most people can relate to common action slips, such as flicking a light switch despite knowledge that the power is out (Cooper & Shallice, 2000). Rules, of course, are much more important when behaviourThe first principle is that the division of a mental program into goals, and then appropriate sub-goals is a fundamental property of successful goal- directed behaviour in humans (Duncan, 2010; Newell & Simon, 1972). Human behaviour is intrinsically goal-oriented; from the moment we wake up, goals begin to guide our behaviour. What we wear today, what we are going to have for breakfast, etc., immediately occupy our awareness. These goals initially exist at an abstract level (Sacerdoti, 1974). For example, when we awake we do not yet possess the final motor commands for preparing breakfast because we have not yet made it to the appropriate environment (the kitchen) and we have not assessed whether we have the right requirements to prepare breakfast. This process of high-level goals being decomposed into appropriate sub-goals to progress towards completion of a supergoal is so fundamental to behaviourthat it is often taken for granted (see Figure 1-1). However, work using cognitive architectures like SOAR and ACT-R has made it clear that this problem of goal/sub-goal management is a fundamental property of how goal- oriented agents behave in the world (Anderson et al., 2004).

When I first read the question for this essay competition, I immediately associated the word 'social' with the word 'personal', i.e. outside of work. Social media is becoming an important source of primary information about the lives of ordinary people like you and me. But I am wary.

INTERSECTING CIRCLES Algorithm and Runtime Form a pointset of all circle centers. Use a WSPD to find the closest pair of centers \(O(n\log n)\). Return disjoint if the distance between these points is > 1, non-disjoint else. Correctedness Disjoint circles \(\iff\) all pairwise circle centers are > 1 dist apart (unit circles, r=1) \(\iff\) the closest pair of points is > 1 dist apart \(\square\).

We address the issue of timeline evaluation and inference on timeline models. We develop an exciting new framework for evaluation, and argue for its theoretical soundness. We also improve upon the state-of-the-art in terms of model inference. We make the first attempts in the literature to apply variational inference methods to the timeline generation problem. In doing so we obtain results that are competitive in terms of the timelines generated with speed-ups of an order of magnitude.

WSPD A) One by definition of the WSPD. B) Depends on decomposition/points chosen. In general it can be \(1\) to \(n-1\). For lower bound, we can make a WSPD of a point that has all pair of points as singleton sets. In our example this WSPD would look like {{a}{b}} {{a}{c}} {{b}{c}}. This is always a valid WSPD for any pointset/s and we have that all pairs appear in the union of some subset-pair once. In terms of a lower bound, consider a WSPD of the form {{a}, {b}}, {{a,b}, {c}} {{a,b}, {d}}, etc. We assume here that this is valid, i.e. all points lie further than sr from the {a,b} circle (which is clearly s * dist(a,b)). In this case a,b appear in n-1 of the unions. C) Given \(s\geq 2\) \(|A|=|B|=1\). Proof is simple and given in lectures. Basically, if p and q are the closest points and \(s\geq 2\) then any point within either of their WSP-sets would have to be closer, a contradiction. D) \(|A|=1\) from above, \(|B| \in[1,n-1]_\). Lower bound when p,q are the closest two points, upper bound when we have {{a}{b,c,...}}.

MODEL We implement a standard averaged perceptron model, as documented in the assignment specification. We also repeated several of the experiments used throughout this analysis with a multinomial Naive Bayes classifier. As we were encouraged to use the Pereptron model and considering the limited amount of space available, the results of this comparison have not been included. It is worth noting that the Naive Bayes model seemed to perform no worse that the Perceptron in terms of accuracy, and often beat it. Furthermore, in terms of model run-time the Naive Bayes was very competitive. VALIDATION DATASET We validated our model on the Internet Advertisement Data Set. We found with a single pass we achieved a 10-fold average accuracy of 0.949. This is reasonably competitive with other benchmarks on this dataset, so we are satisfied with our implementation. (See http://archive.ics.uci.edu/ml/datasets/Internet+Advertisements for benchmarks). The best accuracy (0.951) was found by grid-searching on the number of training iterations -- the best value being three passes through.

RANGE QUERIES For both trees we adopt the convention that to the left child of a node has value strictly smaller, and the right child greater than or equal to. See back of document. (FIG 1): Visited nodes highlighted in yellow. Leafs correspond to points. (FIG 2): The top level data-structure is on x-coordinate. Nodes we visit in the search are highlighted yellow. If we explore an augmented data-structure, we circle the node with red. The internal data-structures are represented as BST arrays. The nodes of the internal data-structure have been highlighted if they are visited during the search.

INTRODUCTION We first introduce the problem and examine several domains where it has been examined. We then address the two primary research questions that arise when comparing different methods: how best to model and evaluate timelines. In each of these two areas we organise the key papers thematically. The contributions of each paper are synthesised into a general discussion about the pros and cons of each approach. The Problem Timeline generation (TLG) is a way of representing a large amount of temporally dependent information concisely. It is query driven; we retrieve a corpus of text linked to some entity, event or other term. The canonical TLG model clusters these articles into topics or stories, selects the most important of these clusters and returns timestamped summaries. It can be seen as a generalisation of the multi-document summarisation task, where we have introduced temporal dependency and structure. CLUSTERING: Our model handles both current and historical articles. Several domains where the TLG model is desirable (e.g. news) are characterised by a large historical catalog and high frequency. Our clustering model has to be scaleable and ideally functional in a streaming context. TOPIC MODELS: Incorporating a topic structure into our model gives us an intuitive and understandable document representation. Looking at the topic distribution also lets us prioritise certain kinds of sentences or stories. TIMELINES: TLG differs from regular multi-document summarisation in its temporal dependence. As such, we value diversity in both the kind of content selected as well as when it was published. SUMMARISATION: While clustering and topic modeling are useful structuring tools, we must still compress the data for human readability. Applications Timeline generation has been applied in a range of domains. Several approaches focus on specific sub-problems in TLG. Althoff et al. timelines entities in the Freebase repository . This knowledge base is comprised of subject and objects linked by predicates; 'Robert Downey Junior' might be linked by 'Starred In' to 'The Avengers. The TLG model seeks to group and select a subset of these subject-predicates that summarise our query entity. Working from knowledge-base instead of a corpus of text provides useful structure. We are spared the task of preprocessing, entity linking, consolidation and data validation on our text corpora. Of course, this relies on the existence of a knowledge base of timestamped facts about our query entity. This is an issue if we want to be able to perform queries on lesser known entities. A similar approach was used by Ahmed et al. on a corpus of academic papers . Single-document variations of the TLG model have been applied to Wikipedia and Wikinews articles . The single-document nature of the task is characterised by a focus on the summarisation and selection tasks over clustering. Timelines took the form of timestamped subsets of the document sentences. A similar approach has been applied to Twitter feeds; Hong et al. applied a single-document version of the TLG model to a user's tweet-history . The most frequent application of TLG models is to a corpus of news articles . A document retrieval service is given a query term; a person, event, company or generic term. The body of documents retrieved form the input for the TLG task. The task then is to group and summarise a potentially large body of articles and provide a compact representation of the query term.

LINE SEGMENTS Algorithm and Runtime - S \(\leftarrow\) Stack - For segment i by ascending x-coordinate: \(O(n)\). - Pop from S until the top element s can be hit by i or S is empty: RTP amortised \(O(1)\). - Emit index of s or 0 in the empty case: \(O(1)\). - Push i onto S: \(O(1)\). To see that our pop-step is amortised \(O(1)\) we note that every element is pushed onto our stack at most one time. We can only pop as many elements as we push, so in total we can pop \(O(n)\) elements. Our loop runs \(O(n)\) times, so the amortised pop-step per iteration is \(O\left({n}\right) = O(1)\). This guarantees overall \(O(n)\) runtime \(\square\). Correctedness 1. If an element is popped at iteration \(i\) it was lower than interval \(i\). If it is lower than interval \(i\), any subsequent interval would have to pass through \(i\) to hit it. As such it cannot be the first-hit point for any subsequent interval. 2. The stack is populated in loop-order. This is ascending x-coord, so higher elements always have larger x-coordinates. When we consider an interval, we keep on throwing away elements until we find one that we can hit. From (2), among all elements in the stack we can hit, this is the one with the largest x-coordinate. From (1), we throw away a value IFF we cannot use it later. As such, there can be no intervals that could possibly hit first that were not present in the stack. The proof is complete.

FAT POINTS Proof by contradiction. Assume that the line segment between points \(P,Q\) has maximum pairwise distance, and that \(Q\) does not lie on a vertex. Let the point on the boundary of our hull found by extending the line \(PQ\) be denoted \(Q'\). This boundary segment is defined between two vertices on our convex hull which we refer to as \(A\) and \(B\). See \(\) for clarification. Q Lies on the Interior of the Hull Clearly \(|PQ'|>|PQ|\). In the following section we show that there is always a line-segment longer than \(|PQ'|\). By the transitive property of inequality this segment must also be longer than \(PQ\), a contradiction. Q Lies on a Hull Edge

ABSTRACT Personal news and content curation is an exciting NLP application. Systems providing this service are often characterised by a collaborative approach that combines human and machine intelligence. As the scope of the problem increases however, so too does the importance of automation. To this end we propose a novel method for scoring news articles and other related content. It is natural to view this problem in a learning-to-rank framework. The training phase of our model first makes use of a pairwise transform. This alters the problem from the ranking of a whole corpus to many individual pairwise comparisons (is article 'a' better than article 'b'). This transformed set is then used to determine the optimal weights in a logistic regression model. These can then be used directly to classify the non-transformed test set. We also perform a comprehensive review and selection process on a large range of candidate features. Our final features involve measures of centrality, informativeness, complexity and within-group similarity.

Kernel-approximation methods in tandem with efficient online classifiers have proven to be a highly effective solution to the occupancy map problem. In this paper we seek to expand upon the work done in this area. Our contributions are twofold. We demonstrate that a Bayesian logistic regression classifier fails to perform better than the more spartan point-estimator/subgradient descent method. Contrastingly, we show that Bayesian optimisation over the hyperparameters of the model is an incredibly powerful and useful tool for this application.

_This is the author’s version of the work. It is posted here by permission of the AAAS for personal use, not for redistribution. The definitive version was published in Science on Vol. 350 no. 6259 pp. 423-426 , DOI: 10.1126/science.aac6933_ Internal stellar magnetic fields are inaccessible to direct observations and little is known about their amplitude, geometry and evolution. We demonstrate that strong magnetic fields in the cores of red giant stars can be identified with asteroseismology. The fields can manifest themselves via depressed dipole stellar oscillation modes, which arises from a magnetic greenhouse effect that scatters and traps oscillation mode energy within the core of the star. The _Kepler_ satellite has observed a few dozen red giants with depressed dipole modes which we interpret as stars with strongly magnetized cores. We find field strengths larger than $\sim\! 10^5 \,{\rm G}$ may produce the observed depression, and in one case we infer a minimum core field strength of $\approx \! \! 10^7 \,{\rm G}$.

The KITP invites collaborating researchers who were participants of selected recent programs to apply to visit, in order to pursue work initiated during their prior KITP visits. The goal is to enhance those collaborative efforts initiated at a KITP program that would clearly benefit from a brief, focused return to the KITP. Groups of 2-6 members with established collaborations arising from the following programs may apply to visit for 1 or 2 weeks during the period October 26-December 18, 2015: - Magnetism, Bad Metals and Superconductivity: Iron Pnictides and Beyond - Galactic Archaeology and Precision Stellar Astrophysics - Dynamics and Evolution of Earth-like Planets - Quantum Gravity Foundations: UV to IR Proposals should be short, consisting of 1-2 pages justifying the group of researchers who will participate, explaining the science and objectives that the group will pursue during the visit, and explicitly stating the preferred dates when the group can come. Visits must be for one or two complete weeks, with all applicants in residence at the same time. Applying to the program consists of two steps: 1. One collaborator must submit the group’s proposal at Form. 2. Each member of the proposed group must also fill out a brief online application at Activities. All application materials must be received by July 5, 2015. Invitations will be issued by July 17, 2015. For those invited to participate, KITP will reserve and directly pay for housing at West Cottages and provide office space in Kohn Hall during the visits. Visitors will be fully responsible for travel, meals and any other expenses.

_This is the author’s version of the work. It is posted here for personal use, not for redistribution. The definitive version was published in Nature on 04 January 2016, DOI:10.1038/nature16171_ Magnetic fields play a role in almost all stages of stellar evolution . Most low-mass stars, including the Sun, show surface fields that are generated by dynamo processes in their convective envelopes . Intermediate-mass stars do not have deep convective envelopes , although 10% exhibit strong surface fields that are presumed to be residuals from the stellar formation process . These stars do have convective cores that might produce internal magnetic fields , and these might even survive into later stages of stellar evolution, but information has been limited by our inability to measure the fields below the stellar surface . Here we use asteroseismology to study the occurrence of strong magnetic fields in the cores of low- and intermediate-mass stars. We have measured the strength of dipolar oscillation modes, which can be suppressed by a strong magnetic field in the core , in over 3,600 red giant stars observed by . About 20% of our sample show mode suppression but this fraction is a strong function of mass. Strong core fields only occur in red giants above 1.1 solar masses (1.1), and the occurrence rate is at least 60% for intermediate-mass stars (1.6–2.0), indicating that powerful dynamos were very common in the convective cores of these stars.

SUMMARY The 2013-2015 Ebola virus disease (EVD) epidemic is caused by the Makona variant of Ebola virus (EBOV). Early in the epidemic, genome sequencing provided insights into virus evolution and transmission, and offered important information for outbreak response. Here we analyze sequences from 232 patients sampled over 7 months in Sierra Leone, along with 86 previously released genomes from earlier in the epidemic. We confirm sustained human-to-human transmission within Sierra Leone and find no evidence for import or export of EBOV across national borders after its initial introduction. Using high-depth replicate sequencing, we observe both host-to-host transmission and recurrent emergence of intrahost genetic variants. We trace the increasing impact of purifying selection in suppressing the accumulation of nonsynonymous mutations over time. Finally, we note changes in the mucin-like domain of EBOV glycoprotein that merit further investigation. These findings clarify the movement of EBOV within the region and describe viral evolution during prolonged human-to-human transmission.

Why do quick brown foxes jump over lazy dogs anyway? From a sociological perspective, it's useful to think about our judgmental attitude to laziness and speed and the way we theorise about sport.