Search for: All records

Abstract The backpropagation method has enabled transformative uses of neural networks. Alternatively, for energy-based models, local learning methods involving only nearby neurons offer benefits in terms of decentralized training, and allow for the possibility of learning in computationally-constrained substrates. One class of local learning methodscontraststhe desired, clamped behavior with spontaneous, free behavior. However, directly contrasting free and clamped behaviors requires explicit memory. Here, we introduce ‘Temporal Contrastive Learning’, an approach that uses integral feedback in each learning degree of freedom to provide a simple form of implicit non-equilibrium memory. During training, free and clamped behaviors are shown in a sawtooth-like protocol over time. When combined with integral feedback dynamics, these alternating temporal protocols generate an implicit memory necessary for comparing free and clamped behaviors, broadening the range of physical and biological systems capable of contrastive learning. Finally, we show that non-equilibrium dissipation improves learning quality and determine a Landauer-like energy cost of contrastive learning through physical dynamics. 

Abstract Retreat of continental ice sheets exposes comminuted sediment in disequilibrium with non-glacial conditions. Weathering of this sediment may create climate feedbacks by altering exchange of greenhouse gases between atmosphere and landscapes. Here we show in a partially deglaciated watershed in southwest Greenland that glacial meltwater contains low concentrations of reactive dissolved organic carbon that enhances weathering of freshly comminuted sediment causing net sequestration of carbon dioxide. In contrast, soil water reactions enhance methanogenesis and carbon dioxide production and create greenhouse gas sources as organic carbon is remineralized. We suggest that a change from greenhouse gas sinks in glacial meltwater to greenhouse gas sources in soil water creates a switch from a negative to positive warming feedback during glacial-interglacial transitions, but a negative warming feedback may return with future anthropogenic warming, glacial retreat, and increased meltwater production. We anticipate changing weathering reactions following exposure also alter nutrient and radiogenic isotope exports. 

in mathematics education, and researchers often turn to advanced natural language processing (NLP) models to analyze classroom dialogues from multiple perspectives. However, utterance-level discourse analysis encounters two primary challenges: (1) multifunctionality, where a single utterance may serve multiple purposes that a single tag cannot capture, and (2) the exclusion of many utterances from domain-specific discourse move classifications, leading to their omission in feedback. To address these challenges, we proposed a multi-perspective discourse analysis that integrates domain-specific talk moves with dialogue act (using the flattened multi-functional SWBD-MASL schema with 43 tags) and discourse relation (applying Segmented Discourse Representation Theory with 16 relations). Our top-down analysis framework enables a comprehensive understanding of utterances that contain talk moves, as well as utterances that do not contain talk moves. This is applied to two mathematics education datasets: TalkMoves (teaching) and SAGA22 (tutoring). Through distributional unigram analysis, sequential talk move analysis, and multi-view deep dive, we discovered meaningful discourse patterns, and revealed the vital role of utterances without talk moves, demonstrating that these utterances, far from being mere fillers, serve crucial functions in guiding, acknowledging, and structuring classroom discourse. These insights underscore the importance of incorporating discourse relations and dialogue acts into AI-assisted education systems to enhance feedback and create more responsive learning environments. Our framework may prove helpful for providing human educator feedback, but also aiding in the development of AI agents that can effectively emulate the roles of both educators and students. 

Abstract Arctic Indigenous food systems are essential to the survival of local communities, but face significant challenges due to environmental, economic, and social pressures. The objective of this study is to elicit values of the mixed Indigenous food system on St. Paul Island, Alaska, and understand their potential as deep leverage points for transformative change in the context of their historical co-evolution with the local food system. To achieve this objective, we engage three generational groups within the Aleut Community of St. Paul Island to obtain local food system stories. From these stories, we (i) elicit historical events that are thought to have shaped the local food system, (ii) identify factors that influence the food system in its present and future states, and (iii) delineate intrinsic, instrumental, and relational food system values. Our findings show that most identified historical events are perceived to have undermined the Indigenous food system and that most factors identified to shape present and future food system states present barriers for community members to engage in traditional practices. Yet, despite this, values that relate to traditional Indigenous livelihoods remain central in the local value system. These results suggest a value change debt, i.e., a time lag between changes in peoples’ held values following changes in the system around them. We propose that this lag provides a window of opportunity to leverage transformative change. We argue that as long as traditional food system values persist, there is potential to reconfigure the food system in a way that embraces these values, enhancing the system's relevance to the community's way of life. 

We present a compendium of HI 21-cm line observations of circumstellar envelopes (CSEs) of 290 evolved stars, mostly (~84%) on the asymptotic giant branch (AGB), made with the 100 m-class, single-dish Nançay Radio Telescope. The observational and data reduction procedures were optimised to separate genuine CSE HI emission from surrounding Galactic line features. For most targets (254), the results have not been previously published. Clear detections were made of 34 objects, for 33 of which the total HI flux and the size of the CSE could be determined. Possible detections were made of 21 objects, and upper limits could be determined for 95 undetected targets, while for 140 objects confusion from Galactic HI emission along the line of sight precluded meaningful upper limits. The collective results of this survey can provide guidance on the detectability of circumstellar HI gas for future mapping and imaging studies. 

Human tutoring interventions play a crucial role in supporting student learning, improving academic performance, and promoting personal growth. This paper focuses on analyzing mathematics tutoring discourse using talk moves—a framework of dialogue acts grounded in Accountable Talk theory. However, scaling the collection, annotation, and analysis of extensive tutoring dialogues to develop machine learning models is a challenging and resource-intensive task. To address this, we present SAGA22, a compact dataset, and explore various modeling strategies, including dialogue context, speaker information, pretraining datasets, and further fine-tuning. By leveraging existing datasets and models designed for classroom teaching, our results demonstrate that supplementary pretraining on classroom data enhances model performance in tutoring settings, particularly when incorporating longer context and speaker information. Additionally, we conduct extensive ablation studies to underscore the challenges in talk move modeling. 

Human tutoring interventions play a crucial role in supporting student learning, improving academic performance, and promoting personal growth. This paper focuses on analyzing mathematics tutoring discourse using talk moves—a framework of dialogue acts grounded in Accountable Talk theory. However, scaling the collection, annotation, and analysis of extensive tutoring dialogues to develop machine learning models is a challenging and resource-intensive task. To address this, we present SAGA22, a compact dataset, and explore various modeling strategies, including dialogue context, speaker information, pretraining datasets, and further fine-tuning. By leveraging existing datasets and models designed for classroom teaching, our results demonstrate that supplementary pretraining on classroom data enhances model performance in tutoring settings, particularly when incorporating longer context and speaker information. Additionally, we conduct extensive ablation studies to underscore the challenges in talk move modeling.