arXiv subjects

James A. Evans

Publications and source records attributed to James A. Evans.

33 records · Page 2Linked to original sources

arXiv2024

Learning from One and Only One Shot

Humans can generalize from only a few examples and from little pretraining on similar tasks. Yet, machine learning (ML) typically requires large data to learn or pre-learn to transfer. Motivated by nativism and artificial general intelligence, we directly model human-innate priors in abstract visual tasks such as character and doodle recognition. This yields a white-box model that learns general-appearance similarity by mimicking how humans naturally ``distort'' an object at first sight. Using just nearest-neighbor classification on this cognitively-inspired similarity space, we achieve human-level recognition with only $1$--$10$ examples per class and no pretraining. This differs from few-shot learning that uses massive pretraining. In the tiny-data regime of MNIST, EMNIST, Omniglot, and QuickDraw benchmarks, we outperform both modern neural networks and classical ML. For unsupervised learning, by learning the non-Euclidean, general-appearance similarity space in a $k$-means style, we achieve multifarious visual realizations of abstract concepts by generating human-intuitive archetypes as cluster centroids.

cs.CV↗

arXiv2024

Limited Diffusion of Scientific Knowledge Forecasts Collapse

Market bubbles emerge when asset prices are driven unsustainably higher than asset values and shifts in belief burst them. We demonstrate the same phenomenon for biomedical knowledge when promising research receives inflated attention. We predict deflationary events by developing a diffusion index that captures whether research areas have been amplified within social and scientific bubbles or have diffused and become evaluated more broadly. We illustrate our diffusion approach contrasting the trajectories of cardiac stem cell research and cancer immunotherapy. We then trace the diffusion of unique 28,504 subfields in biomedicine comprising nearly 1.9M papers and more than 80M citations and demonstrate that limited diffusion of biomedical knowledge anticipates abrupt decreases in popularity. Our analysis emphasizes that restricted diffusion, implying a socio-epistemic bubble, leads to dramatic collapses in relevance and attention accorded to scientific knowledge.

cs.SI↗

arXiv2023

Human Languages with Greater Information Density Increase Communication Speed, but Decrease Conversation Breadth

Human languages vary widely in how they encode information within circumscribed semantic domains (e.g., time, space, color, human body parts and activities), but little is known about the global structure of semantic information and nothing about its relation to human communication. We first show that across a sample of ~1,000 languages, there is broad variation in how densely languages encode information into their words. Second, we show that this language information density is associated with a denser configuration of semantic information. Finally, we trace the relationship between language information density and patterns of communication, showing that informationally denser languages tend toward (1) faster communication, but (2) conceptually narrower conversations within which topics of conversation are discussed at greater depth. These results highlight an important source of variation across the human communicative channel, revealing that the structure of language shapes the nature and texture of human engagement, with consequences for human behavior across levels of society.

cs.CL↗

arXiv2023

Disrupted Routines Anticipate Musical Exploration

Prior research suggests that taste preferences relate to personality traits, values, shifts in mood, and immigration destination, but understanding everyday patterns of listening and the function music plays in life have remained elusive, despite speculations that musical nostalgia may compensate for local disruption. Using more than a hundred million streams of 4 million songs by tens of thousands of international listeners from a global music service catering to local tastes, here we show that breaches in personal routine are systematically associated with personal musical exploration. As people visited new cities and countries, their preferences diversified, converging towards their destinations. As people experienced COVID-19 lock-downs, and then again when they experienced reopenings, their preferences diversified further.

cs.SI↗

arXiv2022

Quantifying hierarchy in scientific teams

This paper provides a detailed description of the data collection and machine learning model used in our recent PNAS paper "Flat Teams Drive Scientific Innovation" Xu et al. [2022a]. Here, we discuss how the features of scientific publication can be used to estimate the implicit hierarchy in the corresponding author teams. Besides, we also describe the method of evaluating the impact of team hierarchy on scientific outputs. More details will be updated in this article continuously. Raw data and Readme document can be accessed in this GitHub repository Xu et al. [2022b].

cs.DL↗

arXiv2021

Policy-Aware Mobility Model Explains the Growth of COVID-19 in Cities

With the continued spread of coronavirus, the task of forecasting distinctive COVID-19 growth curves in different cities, which remain inadequately explained by standard epidemiological models, is critical for medical supply and treatment. Predictions must take into account non-pharmaceutical interventions to slow the spread of coronavirus, including stay-at-home orders, social distancing, quarantine and compulsory mask-wearing, leading to reductions in intra-city mobility and viral transmission. Moreover, recent work associating coronavirus with human mobility and detailed movement data suggest the need to consider urban mobility in disease forecasts. Here we show that by incorporating intra-city mobility and policy adoption into a novel metapopulation SEIR model, we can accurately predict complex COVID-19 growth patterns in U.S. cities ($R^2$ = 0.990). Estimated mobility change due to policy interventions is consistent with empirical observation from Apple Mobility Trends Reports (Pearson's R = 0.872), suggesting the utility of model-based predictions where data are limited. Our model also reproduces urban "superspreading", where a few neighborhoods account for most secondary infections across urban space, arising from uneven neighborhood populations and heightened intra-city churn in popular neighborhoods. Therefore, our model can facilitate location-aware mobility reduction policy that more effectively mitigates disease transmission at similar social cost. Finally, we demonstrate our model can serve as a fine-grained analytic and simulation framework that informs the design of rational non-pharmaceutical interventions policies.

q-bio.PE↗

arXiv2020

Too many cooks: Bayesian inference for coordinating multi-agent collaboration

Collaboration requires agents to coordinate their behavior on the fly, sometimes cooperating to solve a single task together and other times dividing it up into sub-tasks to work on in parallel. Underlying the human ability to collaborate is theory-of-mind, the ability to infer the hidden mental states that drive others to act. Here, we develop Bayesian Delegation, a decentralized multi-agent learning mechanism with these abilities. Bayesian Delegation enables agents to rapidly infer the hidden intentions of others by inverse planning. We test Bayesian Delegation in a suite of multi-agent Markov decision processes inspired by cooking problems. On these tasks, agents with Bayesian Delegation coordinate both their high-level plans (e.g. what sub-task they should work on) and their low-level actions (e.g. avoiding getting in each other's way). In a self-play evaluation, Bayesian Delegation outperforms alternative algorithms. Bayesian Delegation is also a capable ad-hoc collaborator and successfully coordinates with other agent types even in the absence of prior experience. Finally, in a behavioral experiment, we show that Bayesian Delegation makes inferences similar to human observers about the intent of others. Together, these results demonstrate the power of Bayesian Delegation for decentralized multi-agent collaboration.

cs.AI↗

arXiv2020

Human Evaluation of Interpretability: The Case of AI-Generated Music Knowledge

Interpretability of machine learning models has gained more and more attention among researchers in the artificial intelligence (AI) and human-computer interaction (HCI) communities. Most existing work focuses on decision making, whereas we consider knowledge discovery. In particular, we focus on evaluating AI-discovered knowledge/rules in the arts and humanities. From a specific scenario, we present an experimental procedure to collect and assess human-generated verbal interpretations of AI-generated music theory/rules rendered as sophisticated symbolic/numeric objects. Our goal is to reveal both the possibilities and the challenges in such a process of decoding expressive messages from AI sources. We treat this as a first step towards 1) better design of AI representations that are human interpretable and 2) a general methodology to evaluate interpretability of AI-discovered knowledge representations.

cs.HC↗

arXiv2019

Quantifying dynamics of failure across science, startups, and security

Human achievements are often preceded by repeated attempts that initially fail, yet little is known about the mechanisms governing the dynamics of failure. Here, building on the rich literature on innovation, human dynamics and learning, we develop a simple one-parameter model that mimics how successful future attempts build on those past. Analytically solving this model reveals a phase transition that separates dynamics of failure into regions of stagnation or progression, predicting that near the critical threshold, agents who share similar characteristics and learning strategies may experience fundamentally different outcomes following failures. Below the critical point, we see those who explore disjoint opportunities without a pattern of improvement, and above it, those who exploit incremental refinements to systematically advance toward success. The model makes several empirically testable predictions, demonstrating that those who eventually succeed and those who do not may be initially similar, yet are characterized by fundamentally distinct failure dynamics in terms of the efficiency and quality of each subsequent attempt. We collected large-scale data from three disparate domains, tracing repeated attempts by (i) NIH investigators to fund their research, (ii) innovators to successfully exit their startup ventures, and (iii) terrorist organizations to post casualties in violent attacks, finding broadly consistent empirical support across all three domains. Together, our findings unveil identifiable yet previously unknown early signals that allow us to identify failure dynamics that will lead to ultimate victory or defeat. Given the ubiquitous nature of failures and the paucity of quantitative approaches to understand them, these results represent a crucial step toward deeper understanding of the complex dynamics beneath failures, the essential prerequisites for success.

physics.soc-ph↗

arXiv2018

Homogenisation of thin periodic frameworks with high-contrast inclusions

We analyse a problem of two-dimensional linearised elasticity for a two-component periodic composite, where one of the components consists of disjoint soft inclusions embedded in a rigid framework. We consider the case when the contrast between the elastic properties of the framework and the inclusions, as well as the ratio between the period of the composite and the framework thickness increase as the period of the composite becomes smaller. We show that in this regime the elastic displacement converges to the solution of a special two-scale homogenised problem, where the microscopic displacement of the framework is coupled both to the slowly-varying "macroscopic" part of the solution and to the displacement of the inclusions. We prove the convergence of the spectra of the corresponding elasticity operators to the spectrum of the homogenised operator with a band-gap structure.

math-ph↗

arXiv2018

The Geometry of Culture: Analyzing Meaning through Word Embeddings

We demonstrate the utility of a new methodological tool, neural-network word embedding models, for large-scale text analysis, revealing how these models produce richer insights into cultural associations and categories than possible with prior methods. Word embeddings represent semantic relations between words as geometric relationships between vectors in a high-dimensional space, operationalizing a relational model of meaning consistent with contemporary theories of identity and culture. We show that dimensions induced by word differences (e.g. man - woman, rich - poor, black - white, liberal - conservative) in these vector spaces closely correspond to dimensions of cultural meaning, and the projection of words onto these dimensions reflects widely shared cultural connotations when compared to surveyed responses and labeled historical data. We pilot a method for testing the stability of these associations, then demonstrate applications of word embeddings for macro-cultural investigation with a longitudinal analysis of the coevolution of gender and class associations in the United States over the 20th century and a comparative analysis of historic distinctions between markers of gender and class in the U.S. and Britain. We argue that the success of these high-dimensional models motivates a move towards "high-dimensional theorizing" of meanings, identities and cultural processes.

cs.CL↗

arXiv2018

Centralized "big science" communities more likely generate non-replicable results

Growing concern that most published results, including those widely agreed upon, may be false are rarely examined against rapidly expanding research production. Replications have only occurred on small scales due to prohibitive expense and limited professional incentive. We introduce a novel, high-throughput replication strategy aligning 51,292 published claims about drug-gene interactions with high-throughput experiments performed through the NIH LINCS L1000 program. We show (1) that unique claims replicate 19% more frequently than at random, while those widely agreed upon replicate 45% more frequently, manifesting collective correction mechanisms in science; but (2) centralized scientific communities perpetuate claims that are less likely to replicate even if widely agreed upon, demonstrating how centralized, overlapping collaborations weaken collective understanding. Decentralized research communities involve more independent teams and use more diverse methodologies, generating the most robust, replicable results. Our findings highlight the importance of science policies that foster decentralized collaboration to promote robust scientific advance.

cs.SI↗

arXiv2017

Large Teams Have Developed Science and Technology; Small Teams Have Disrupted It

Teams dominate the production of high-impact science and technology. Analyzing teamwork from more than 50 million papers, patents, and software products, 1954-2014, we demonstrate across this period that larger teams developed recent, popular ideas, while small teams disrupted the system by drawing on older and less prevalent ideas. Attention to work from large teams came immediately, while advances by small teams succeeded further into the future. Differences between small and large teams magnify with impact - small teams have become known for disruptive work and large teams for developing work. Differences in topic and re- search design account for part of the relationship between team size and disruption, but most of the effect occurs within people, controlling for detailed subject and article type. Our findings suggest the importance of supporting both small and large teams for the sustainable vitality of science and technology.

physics.soc-ph↗

arXiv2015

Full two-scale asymptotic expansion and higher-order constitutive laws in the homogenisation of the system of Maxwell equations

For the system of Maxwell equations of electromagnetism in an $l$-periodic composite medium of overall size $L$ ($0<l<L<\infty$), in the low-frequency quasistatic approximation, we develop an electromagnetic version of strain-gradient theories, where the magnetic field is not a function of the magnetic induction alone but also of its spatial gradients, and the electric field depends not only on the displacement but also on displacement gradients. Following the work (Smyshlyaev, V.P., Cherednichenko, K.D., 2000. On rigorous derivation of strain gradient effects in the overall behaviour of periodic heterogeneous media, ${\mathit J.\ Mech.\ Phys.\ Solids\ }{\mathbf{48}},$ $1325-1357$), we develop a combination of variational and asymptotic approaches to the multiscale analysis of the Maxwell system. We provide rigorous convergence estimates of higher order of smallness with respect to the inverse of the "scale separation parameter" $L/l.$ Using a special "ensemble averaging" procedure for a family of periodic problems, we derive an infinite-order version of the classical homogenised system of Maxwell equations.

math-ph↗

arXiv2013

Tradition and Innovation in Scientists' Research Strategies

What factors affect a scientist's choice of research problem? Qualitative research in the history, philosophy, and sociology of science suggests that this choice is shaped by an "essential tension" between the professional demand for productivity and a conflicting drive toward risky innovation. We examine this tension empirically in the context of biomedical chemistry. We use complex networks to represent the evolving state of scientific knowledge, as expressed in publications. We then define research strategies relative to these networks. Scientists can introduce novel chemicals or chemical relationships--or delve deeper into known ones. They can consolidate existing knowledge clusters, or bridge distant ones. Analyzing such choices in aggregate, we find that the distribution of strategies remains remarkably stable, even as chemical knowledge grows dramatically. High-risk strategies, which explore new chemical relationships, are less prevalent in the literature, reflecting a growing focus on established knowledge at the expense of new opportunities. Research following a risky strategy is more likely to be ignored but also more likely to achieve high impact and recognition. While the outcome of a risky strategy has a higher expected reward than the outcome of a conservative strategy, the additional reward is insufficient to compensate for the additional risk. By studying the winners of 137 different prizes in biomedicine and chemistry, we show that the occasional "gamble" for extraordinary impact is the most plausible explanation for observed levels of risk-taking. Our empirical demonstration and unpacking of the "essential tension" suggests policy interventions that may foster more innovative research.

physics.soc-ph↗