skip to main content

Title: Learning to communicate about shared procedural abstractions
Many real-world tasks require agents to coordinate their behavior to achieve shared goals. Successful collaboration requires not only adopting the same communicative conventions, but also grounding these conventions in the same task-appropriate conceptual abstractions. We investigate how humans use natural language to collaboratively solve physical assembly problems more effectively over time. Human participants were paired up in an online environment to reconstruct scenes containing two block towers. One participant could see the target towers, and sent assembly instructions for the other participant to reconstruct. Participants provided increasingly concise instructions across repeated attempts on each pair of towers, using more abstract referring expressions that captured each scene's hierarchical structure. To explain these findings, we extend recent probabilistic models of ad hoc convention formation with an explicit perceptual learning mechanism. These results shed light on the inductive biases that enable intelligent agents to coordinate upon shared procedural abstractions.
; ; ; ;
Award ID(s):
Publication Date:
Journal Name:
Proceedings of the Annual Conference of the Cognitive Science Society
Page Range or eLocation-ID:
Sponsoring Org:
National Science Foundation
More Like this
  1. The DeepLearningEpilepsyDetectionChallenge: design, implementation, andtestofanewcrowd-sourced AIchallengeecosystem Isabell Kiral*, Subhrajit Roy*, Todd Mummert*, Alan Braz*, Jason Tsay, Jianbin Tang, Umar Asif, Thomas Schaffter, Eren Mehmet, The IBM Epilepsy Consortium◊ , Joseph Picone, Iyad Obeid, Bruno De Assis Marques, Stefan Maetschke, Rania Khalaf†, Michal Rosen-Zvi† , Gustavo Stolovitzky† , Mahtab Mirmomeni† , Stefan Harrer† * These authors contributed equally to this work † Corresponding authors:,,,, ◊ Members of the IBM Epilepsy Consortium are listed in the Acknowledgements section J. Picone and I. Obeid are with Temple University, USA. T. Schaffter is with Sage Bionetworks, USA. E. Mehmet is with the University of Illinois at Urbana-Champaign, USA. All other authors are with IBM Research in USA, Israel and Australia. Introduction This decade has seen an ever-growing number of scientific fields benefitting from the advances in machine learning technology and tooling. More recently, this trend reached the medical domain, with applications reaching from cancer diagnosis [1] to the development of brain-machine-interfaces [2]. While Kaggle has pioneered the crowd-sourcing of machine learning challenges to incentivise data scientists from around the world to advance algorithm and model design, the increasing complexity of problem statements demands of participants to be expert datamore »scientists, deeply knowledgeable in at least one other scientific domain, and competent software engineers with access to large compute resources. People who match this description are few and far between, unfortunately leading to a shrinking pool of possible participants and a loss of experts dedicating their time to solving important problems. Participation is even further restricted in the context of any challenge run on confidential use cases or with sensitive data. Recently, we designed and ran a deep learning challenge to crowd-source the development of an automated labelling system for brain recordings, aiming to advance epilepsy research. A focus of this challenge, run internally in IBM, was the development of a platform that lowers the barrier of entry and therefore mitigates the risk of excluding interested parties from participating. The challenge: enabling wide participation With the goal to run a challenge that mobilises the largest possible pool of participants from IBM (global), we designed a use case around previous work in epileptic seizure prediction [3]. In this “Deep Learning Epilepsy Detection Challenge”, participants were asked to develop an automatic labelling system to reduce the time a clinician would need to diagnose patients with epilepsy. Labelled training and blind validation data for the challenge were generously provided by Temple University Hospital (TUH) [4]. TUH also devised a novel scoring metric for the detection of seizures that was used as basis for algorithm evaluation [5]. In order to provide an experience with a low barrier of entry, we designed a generalisable challenge platform under the following principles: 1. No participant should need to have in-depth knowledge of the specific domain. (i.e. no participant should need to be a neuroscientist or epileptologist.) 2. No participant should need to be an expert data scientist. 3. No participant should need more than basic programming knowledge. (i.e. no participant should need to learn how to process fringe data formats and stream data efficiently.) 4. No participant should need to provide their own computing resources. In addition to the above, our platform should further • guide participants through the entire process from sign-up to model submission, • facilitate collaboration, and • provide instant feedback to the participants through data visualisation and intermediate online leaderboards. The platform The architecture of the platform that was designed and developed is shown in Figure 1. The entire system consists of a number of interacting components. (1) A web portal serves as the entry point to challenge participation, providing challenge information, such as timelines and challenge rules, and scientific background. The portal also facilitated the formation of teams and provided participants with an intermediate leaderboard of submitted results and a final leaderboard at the end of the challenge. (2) IBM Watson Studio [6] is the umbrella term for a number of services offered by IBM. Upon creation of a user account through the web portal, an IBM Watson Studio account was automatically created for each participant that allowed users access to IBM's Data Science Experience (DSX), the analytics engine Watson Machine Learning (WML), and IBM's Cloud Object Storage (COS) [7], all of which will be described in more detail in further sections. (3) The user interface and starter kit were hosted on IBM's Data Science Experience platform (DSX) and formed the main component for designing and testing models during the challenge. DSX allows for real-time collaboration on shared notebooks between team members. A starter kit in the form of a Python notebook, supporting the popular deep learning libraries TensorFLow [8] and PyTorch [9], was provided to all teams to guide them through the challenge process. Upon instantiation, the starter kit loaded necessary python libraries and custom functions for the invisible integration with COS and WML. In dedicated spots in the notebook, participants could write custom pre-processing code, machine learning models, and post-processing algorithms. The starter kit provided instant feedback about participants' custom routines through data visualisations. Using the notebook only, teams were able to run the code on WML, making use of a compute cluster of IBM's resources. The starter kit also enabled submission of the final code to a data storage to which only the challenge team had access. (4) Watson Machine Learning provided access to shared compute resources (GPUs). Code was bundled up automatically in the starter kit and deployed to and run on WML. WML in turn had access to shared storage from which it requested recorded data and to which it stored the participant's code and trained models. (5) IBM's Cloud Object Storage held the data for this challenge. Using the starter kit, participants could investigate their results as well as data samples in order to better design custom algorithms. (6) Utility Functions were loaded into the starter kit at instantiation. This set of functions included code to pre-process data into a more common format, to optimise streaming through the use of the NutsFlow and NutsML libraries [10], and to provide seamless access to the all IBM services used. Not captured in the diagram is the final code evaluation, which was conducted in an automated way as soon as code was submitted though the starter kit, minimising the burden on the challenge organising team. Figure 1: High-level architecture of the challenge platform Measuring success The competitive phase of the "Deep Learning Epilepsy Detection Challenge" ran for 6 months. Twenty-five teams, with a total number of 87 scientists and software engineers from 14 global locations participated. All participants made use of the starter kit we provided and ran algorithms on IBM's infrastructure WML. Seven teams persisted until the end of the challenge and submitted final solutions. The best performing solutions reached seizure detection performances which allow to reduce hundred-fold the time eliptologists need to annotate continuous EEG recordings. Thus, we expect the developed algorithms to aid in the diagnosis of epilepsy by significantly shortening manual labelling time. Detailed results are currently in preparation for publication. Equally important to solving the scientific challenge, however, was to understand whether we managed to encourage participation from non-expert data scientists. Figure 2: Primary occupation as reported by challenge participants Out of the 40 participants for whom we have occupational information, 23 reported Data Science or AI as their main job description, 11 reported being a Software Engineer, and 2 people had expertise in Neuroscience. Figure 2 shows that participants had a variety of specialisations, including some that are in no way related to data science, software engineering, or neuroscience. No participant had deep knowledge and experience in data science, software engineering and neuroscience. Conclusion Given the growing complexity of data science problems and increasing dataset sizes, in order to solve these problems, it is imperative to enable collaboration between people with differences in expertise with a focus on inclusiveness and having a low barrier of entry. We designed, implemented, and tested a challenge platform to address exactly this. Using our platform, we ran a deep-learning challenge for epileptic seizure detection. 87 IBM employees from several business units including but not limited to IBM Research with a variety of skills, including sales and design, participated in this highly technical challenge.« less
  2. The ACM/IEEE CS 2013 report recommends fifteen hours of parallel & distributed computing (PDC) education for every undergraduate. This workshop illustrates the use of the Raspberry Pi as an inexpensive, multicore platform for teaching shared-memory parallel programming. The inexpensive and tactile nature of the Raspberry Pi enables each student to experience her own parallel multiprocessor through sight and touch. In this hands-on workshop, we will teach attendees how they can leverage the Raspberry Pi and the OpenMP library to teach shared-memory parallel concepts in their own classrooms. All CS educators who are interested in learning about the Raspberry Pi, shared memory parallelism, and OpenMP are encouraged to attend. In Part I of the workshop, each participant will connect to and learn about the Raspberry Pi's multicore capabilities. In Part II, each participant will engage in self-paced, hands-on exploration of basic parallel computing concepts using the OpenMP "patternlets" from In Part III, participants will investigate more complex applications, such as numeric integration and drug design and study how these applications can be parallelized using OpenMP. We will conclude the workshop with a series of lightning talks discussing how the Raspberry Pi has been used to teach parallel computing concepts atmore »different institutions. We will also present a summary of student perceptions of the Raspberry Pi. All materials from this workshop will be freely available from Space is limited to 20 participants. A laptop is required.« less
  3. Abstract Background To compare the performance (as determined by lower extremity kinematics) of knee exercises in healthy middle-aged and older individuals immediately after instruction and one week later. Methods This is a cross-sectional study in a laboratory setting. Nineteen healthy volunteers (age [y] 63.1 ± 8.6, mass [kg] 76.3 ± 14.7, height [m] 1.7 ± 0.1) participated in this study. High speed video and reflective markers were used to track motion during four exercises. The exercises were knee flexion, straight leg raise, and “V “in supine position, and hip abduction in side lying position. All participants received verbal and tactile cues during the training phase and the therapist observed and, if necessary, corrected the exercises. Upon return a week later the participants performed the same exercises without any further instructions. Knee and hip sagittal and rotational angles were extracted from the motion capture. A repeated measures t-test was used to compare the motions between two visits. Results Participants demonstrated more knee flexion during straight leg raise and “V in” exercises at the 2nd visit compared to the 1st visit (both p  <  0.05). During the “V out” exercise, they performed more external rotation ( p  <  0.05) while they showed more internal rotation during the “V in” exercisemore »at the 2nd visit compared to the 1st visit. Conclusions Exercise performance declined significantly in healthy middle-aged and older individuals one week after instruction. This decline occurred despite an instructional exercise sheet being given to every participant. Other approaches designed to help individuals retain the ability to perform rehabilitative exercises correctly need to be explored.« less
  4. Abstract

    We present CELER (Corpus of Eye Movements in L1 and L2 English Reading), a broad coverage eye-tracking corpus for English. CELER comprises over 320,000 words, and eye-tracking data from 365 participants. Sixty-nine participants are L1 (first language) speakers, and 296 are L2 (second language) speakers from a wide range of English proficiency levels and five different native language backgrounds. As such, CELER has an order of magnitude more L2 participants than any currently available eye movements dataset with L2 readers. Each participant in CELER reads 156 newswire sentences from the Wall Street Journal (WSJ), in a new experimental design where half of the sentences are shared across participants and half are unique to each participant. We provide analyses that compare L1 and L2 participants with respect to standard reading time measures, as well as the effects of frequency, surprisal, and word length on reading times. These analyses validate the corpus and demonstrate some of its strengths. We envision CELER to enable new types of research on language processing and acquisition, and to facilitate interactions between psycholinguistics and natural language processing (NLP).

  5. Corporate involvement in open source software (OSS) communities has increased substantially in recent years. Often this takes the form of company employees devoting their time to contribute code to the efforts of projects in these communities. Ideology has traditionally served to motivate, coordinate, and guide volunteer contributions to OSS communities. As employees represent an increasing proportion of the participants in OSS communities, the role of OSS ideology in guiding their commitment and code contributions is unknown. In this research, we argue that OSS ideology misfit has important implications for companies and the OSS communities to which their employees contribute, since their engagement in such communities is not necessarily voluntary. We conceptualize two different types of misfit: OSS ideology under-fit, whereby an employee embraces an OSS ideology more than their coworkers or OSS community do, and OSS ideology overfit, whereby an employee perceives that their coworkers or OSS community embrace the OSS ideology more strongly than the employee does. To develop a set of hypotheses about the implications of these two types of misfit for employee commitment to the company and commitment to the OSS community, we draw on selfdetermination theory. We test the hypotheses in a field study of 186more »employees who participate in an OSS community. We find that OSS ideology under-fit impacts the company and the community in the same way: it decreases employee commitment to the company and commitment to the OSS community. In contrast, we find that OSS ideology over-fit increases commitment to the company but decreases commitment to the OSS community. Finally, we find that employees’ commitment to their company reinforces the impact of their commitment to the OSS community in driving ongoing code contributions. This provides a holistic view of OSS ideology and its impacts among an increasingly pervasive yet understudied type of participant in OSS research. It provides insights for companies that are considering assigning their employees to work in OSS communities as well as for OSS communities that are partnering with these companies.« less