Browse Articles
Discover research articles across all indexed journals
Krylov diagonalization of large many-body Hamiltonians on a quantum processor
Evolving general cooperation with a Bayesian theory of mind
Theories of the evolution of cooperation through reciprocity explain how unrelated self-interested individuals can accomplish more together than they can on their own. The most prominent theories of reciprocity, such as tit-for-tat or win-stay-lose-shift, are inflexible automata that lack a theory of mind—the human ability to infer the hidden mental states in others’ minds. Here, we develop a model of reciprocity with a theory of mind, the Bayesian Reciprocator. When making decisions, this model does not simply seek to maximize its own payoff. Instead, it also values the payoffs of others—but only to the extent it believes that those others are also cooperating in the same way. To compute its beliefs about others, the Bayesian Reciprocator uses a probabilistic and generative approach to infer the latent preferences, beliefs, and strategies of others through interaction and observation. We evaluate the Bayesian Reciprocator using a generator over games where every interaction is unique, as well as in classic environments such as the iterated prisoner’s dilemma. The Bayesian Reciprocator enables the emergence of both direct-reciprocity when games are repeated and indirect-reciprocity when interactions are one-shot but observable to others. In an evolutionary competition, the Bayesian Reciprocator outcompetes existing automata strategies and sustains cooperation across a larger range of environments and noise settings than prior approaches. This work quantifies the advantage of a theory of mind for cooperation in an evolutionary game theoretic framework and suggests avenues for building artificially intelligent agents with more human-like learning mechanisms that can cooperate across many environments.
Apolipoprotein M attenuates age-related macular degeneration phenotypes via sphingosine-1-phosphate signaling and lysosomal lipid catabolism
Ancient DNA reveals diverse community organizations in the 5th millennium BCE Carpathian Basin
Abstract Little is known about the genetic connection system and community organization of Late Neolithic and Early Copper Age populations of the Carpathian Basin. Here, we present a comprehensive genetic investigation of these populations, leveraging whole genome data from 125 individuals. Using population genetics, kinship analyses and the study of networks of identity-by-descent haplotype segment sharing, we elucidate the social and genetic dynamics of these communities between 4800−3900 calibrated years BCE. Despite changes in settlement patterns, burial practices, and material culture, we document a high degree of genetic continuity. While one set of individuals from a large community cemetery is genetically diverse, another site is more homogenous and closed, with numerous consanguineous relationships and evidence of patrilineality and patrilocality. In this work, we document important differences in kinship systems in contemporaneous Early Copper Age communities using similar material culture and living only about 100 km apart.
Decreased mitochondrial activity in the demyelinating cerebellum of progressive multiple sclerosis and chronic EAE contributes to Purkinje cell loss
In multiple sclerosis (MS), cerebellar gray matter atrophy, white matter demyelination, and Purkinje cell (PC) loss have been linked to tremors, impaired motor control, and loss of coordination. Similar pathologies have been observed in the mouse model of MS, experimental autoimmune encephalomyelitis (EAE). This study hypothesized that inflammatory demyelination of the cerebellum alters overall mitochondrial function and is a contributor to axon degeneration and PC loss. Postmortem cerebellar tissue from MS patients, particularly those with secondary progressive MS, showed decreased mitochondrial complex IV (COXIV) activity and significant PC loss. Inflammation, PC axon demyelination, axon degeneration, and parallel fiber loss were also evident. These findings were mirrored in late-stage EAE mice, which also showed increased inflammation and demyelination, reduced PC COXIV activity, and overall PC loss. Further analysis of EAE mice revealed altered mitochondrial structure, modified mitochondrial respiration, and reduced levels of mitochondrial genes involved in energy production. These findings indicate that both human MS and mouse EAE share similar cerebellar changes linked to mitochondrial dysfunction. Thus, late-stage EAE is a valuable model for studying MS-related cerebellar pathology, and mitochondria may be a potential therapeutic target for MS treatment.
Quantum spin Hall effect in magnetic graphene
A distinct LHCI arrangement is recruited to photosystem I in Fe-starved green algae
Iron (Fe) availability limits photosynthesis at a global scale where Fe-rich photosystem (PS) I abundance is drastically reduced in Fe-poor environments. We used single-particle cryoelectron microscopy to reveal a unique Fe starvation-dependent arrangement of light-harvesting chlorophyll (LHC) proteins where Fe starvation–induced TIDI1 is found in an additional tetramer of LHC proteins associated with PSI in Dunaliella tertiolecta and Dunaliella salina . These cosmopolitan green algae are resilient to poor Fe nutrition. TIDI1 is a distinct LHC protein that co-occurs in diverse algae with flavodoxin (an Fe-independent replacement for the Fe-containing ferredoxin). The antenna expansion in eukaryotic algae we describe here is reminiscent of the iron-starvation induced ( isiA- encoding) antenna ring in cyanobacteria, which typically co-occurs with isiB , encoding flavodoxin. Our work showcases the convergent strategies that evolved after the Great Oxidation Event to maintain PSI capacity.
In vivo self-renewal and expansion of quiescent stem cells from a non-human primate
Abstract The development of non-human primate models is essential for the fields of developmental and regenerative biology because those models will more closely approximate human biology than do murine models. Based on single cell RNAseq and fluorescence-activated cell sorting, we report the identification and functional characterization of two quiescent stem cell populations (skeletal muscle stem cells (MuSCs) and mesenchymal stem cells termed fibro-adipogenic progenitors (FAPs)) in the non-human primate Microcebus murinus (the gray mouse lemur). We demonstrate in vivo proliferation, differentiation, and self-renewal of both MuSCs and FAPs. By combining cell phenotyping with cross-species molecular profiling and pharmacological interventions, we show that mouse lemur MuSCs and FAPs are more similar to human than to mouse counterparts. We identify unexpected gene targets involved in regulating primate MuSC proliferation and primate FAP adipogenic differentiation. Moreover, we find that the cellular composition of mouse lemur muscle better models human muscle than does macaque ( Macaca fascicularis ) muscle. Finally, we note that our approach presents as a generalizable pipeline for the identification, isolation, and characterization of stem cell populations in new animal models.
Increased sea-level contribution from northwestern Greenland for models that reproduce observations
State-of-the-art ice sheet model simulations used in the Ice Sheet Model Intercomparison Project (ISMIP) that informs the Intergovernmental Panel on Climate Change tend to underestimate observed mass loss from the Greenland Ice Sheet, leading to the question of whether future sea-level rise may be larger than projected. We use one of these models, the Ice-sheet and Sea-level System Model, to investigate how transient calibration impacts historical and projection simulations. Transient calibration is an emerging capability in ice flow models; it uses time series of surface observations and time-dependent physics to constrain uncertain model parameters—in this case, the basal friction coefficient in the sliding law. With more constraints than the common snapshot inversion method, transient calibration has been shown to better capture trends in ice dynamics. Here, we apply both methods to northwestern Greenland, a region undergoing rapid changes. For simulations initialized with the snapshot inversion, we find that subsequent modeled velocities are generally too slow, leading to an underestimation of the mass loss. With transient calibration, however, our simulation better matches a time series of observed velocities, bringing it within observational error for mass loss; however, the fit to observed surface elevation is slightly reduced. Together with the ISMIP results, our simulations show that reproducing the high rates of historical mass loss leads to greater projected sea-level contribution from this region over the coming century. Finally, we suggest a path forward for making transient calibration scalable to the entire Greenland Ice Sheet.
Ralph Holloway (1935–2025): Pioneering paleoneurologist
Ralph Holloway pioneered and developed the field of hominin paleoneurology. Although Holloway’s undergraduate degrees were in metallurgical engineering and geology, at graduate school his interests switched to the brain. Holloway and his graduate students explored many aspects of the macro- and microstructure of the brain of extant primates but he focused on the recent evolutionary history of the human brain. The infant skull of the first African early hominin discovered at Taungs by Raymond Dart included a natural brain endocast. Natural endocasts as well-preserved as that of the Taungs infant are rare, but Holloway reasoned that if a way could be found to reproduce the endocranial morphology of early hominin crania then he would be able to track brain evolution within the human clade. Holloway realized that if liquid latex was introduced into the cranial cavity and left to cure, it could be extracted via the foramen magnum and then be used to make a facsimile of that individual’s brain. Modern imaging methods have made Holloway’s technique redundant, but for many years, it was the only way to access the endocranial morphology of fossil hominins.
Collective artificial intelligence and evolutionary dynamics
Unilateral incentive alignment in two-agent stochastic games
Multiagent learning is challenging when agents face mixed-motivation interactions, where conflicts of interest arise as agents independently try to optimize their respective outcomes. Recent advancements in evolutionary game theory have identified a class of “zero-determinant” strategies, which confer an agent with significant unilateral control over outcomes in repeated games. Building on these insights, we present a comprehensive generalization of zero-determinant strategies to stochastic games, encompassing dynamic environments. We propose an algorithm that allows an agent to discover strategies enforcing predetermined linear (or approximately linear) payoff relationships. Of particular interest is the relationship in which both payoffs are equal, which serves as a proxy for fairness in symmetric games. We demonstrate that an agent can discover strategies enforcing such relationships through experience alone, without coordinating with an opponent. In finding and using such a strategy, an agent (“enforcer”) can incentivize optimal and equitable outcomes, circumventing potential exploitation. In particular, from the opponent’s viewpoint, the enforcer transforms a mixed-motivation problem into a cooperative problem, paving the way for more collaboration and fairness in multiagent systems.
Picking strategies in games of cooperation
Evolutionary game theory (EGT) has been pivotal in the study of cooperation, offering formal models that account for how cooperation may arise in groups of selfish, but simple agents. This is done by inspecting the complex dynamics arising from simple interactions between a few strategies in a large population. As such, the strategies at stake are typically hand-picked by the modeler, resulting in a system with many more individuals in the population than strategies available to them. In the presence of noise and with multiple equilibria, the choice of strategies can considerably alter the emergent dynamics. As a result, model outcomes may not be robust to how the strategy set is chosen, sometimes misrepresenting the conditions required for cooperation to emerge. We propose three principles that can lead to a more systematic choice of the strategies in EGT models of cooperation. These are the inclusion of all computationally equivalent strategies; explicit microeconomic models of interactions, and a connection between stylized facts and model assumptions. Further, we argue that new methods arising in AI may offer a promising path toward richer models. These richer models can push the field of cooperation forward together with the principles described above. At the same time, AI may benefit from connecting to the more abstract models of EGT. We provide and discuss examples to substantiate these claims.
European mammal turnover driven by a global rapid warming event preceding the Paleocene–Eocene Thermal Maximum
A brief global warming event known as the Pre-Onset Excursion (POE) occurred just before the Paleocene–Eocene Thermal Maximum (PETM, 56 Mya). The deconvolution of the evolutionary consequences of these two hyperthermal events is puzzling because of their close temporal proximity and the lack of comprehensive, well-calibrated paleontological records, especially in terrestrial environments. As a consequence, the impact of the POE on mammalian evolution and its role in shaping PETM faunas remains unclear. Here, we report from France a mammalian fauna, named Albas, which is interpreted to postdate the POE and predate the PETM. The absence of artiodactyls, perissodactyls, and euprimates at Albas lends support to the controversial hypothesis that these “modern” mammal groups appeared in the European fossil record during the PETM. In contrast, Albas yielded the European first definitive Paleocene record of metatherians, paromomyid primates, “creodonts,” and rodents, challenging the assumption that these groups migrated into Europe during the PETM. Because the majority of them originated from North American pre-POE species, we tentatively suggest that these “precursor” dispersers entered Europe during the POE. Similar to the modern orders during the PETM, these “precursor” dispersers likely entered Europe through corridors in the continuous evergreen forest belt at high latitudes. Our findings highlight how a brief warming event in the Arctic during the latest Paleocene, such as the POE (which could result in a release of carbon into the atmosphere similar to cumulative ongoing anthropogenic emissions), significantly influenced the evolutionary dynamics of European mammals.
Perceptual interventions ameliorate statistical discrimination in learning agents
Choosing social partners is a potentially demanding task which involves paying attention to the right information while disregarding salient but possibly irrelevant features. The resultant trade-off between cost of evaluation and quality of decisions can lead to undesired bias. Information-processing abilities mediate this trade-off, where individuals with higher ability choose better partners leading to higher performance. By altering the salience of features, technology can modulate the effect of information-processing limits, potentially increasing or decreasing undesired biases. Here, we use game theory and multiagent reinforcement learning to investigate how undesired biases emerge, and how a technological layer (in the form of a perceptual intervention) between individuals and their environment can ameliorate such biases. Our results show that a perceptual intervention designed to increase the salience of outcome-relevant features can reduce bias in agents making partner choice decisions. Individuals learning with a perceptual intervention showed less bias due to decreased reliance on features that only spuriously correlate with behavior. Mechanistically, the perceptual intervention effectively increased the information-processing abilities of the individuals. Our results highlight the benefit of using multiagent reinforcement learning to model theoretically grounded social behaviors, particularly when real-world complexity prohibits fully analytical approaches.
Mapping encounters between Antarctic krill fishing vessels and air-breathing krill predators using acoustic data from the fishery
Antarctic krill is a keystone species in the Antarctic marine ecosystem and the target of a growing fishery. Given the ecological importance of krill, concerns have been raised about potential negative impacts of fishing on the Southern Ocean ecosystem. Resource-efficient approaches to fisheries monitoring are particularly valuable in this context due to the high costs associated with data collection in Antarctica. In this study, we trained a segmentation model (U-Net) to extract dives of air-breathing krill predators from more than 30,000 h of active acoustic data collected by three krill fishing vessels over six years. We were able to characterize the temporal and spatial dynamics of predator-vessel co-occurrences, which aligned well with the findings from more costly tracking studies. For example, we found that encounters with whales consistently peaked in autumn around the Antarctic Peninsula, when whales are building up fat reserves for their migration to breeding grounds. We also demonstrated that protection measures, introduced to protect breeding penguins at the Antarctic Peninsula, have simply shifted penguin-vessel encounters to the South Orkney Islands, where the affected colonies are not currently monitored. Our approach, results, and application example demonstrate how acoustic data from fishing vessels can provide important information to support fisheries management. As a by-product of fishing operations, these data are cost-effective, offering unique temporal and spatial coverage and providing a useful basis for rapid, low-level assessments of the fishery’s interaction with the wider ecosystem. This is particularly important given the unpredictable dynamics of krill fishery management decision-making.
Republicans are flagged more often than Democrats for sharing misinformation on X’s Community Notes
We use crowd-sourced assessments from X’s Community Notes program to examine whether there are partisan differences in the sharing of misleading information. Unlike previous studies, misleadingness here is determined by agreement across a diverse community of platform users, rather than by fact-checkers. We find that 2.3 times more posts by Republicans are flagged as misleading compared to posts by Democrats. These results are not base rate artifacts, as we find no meaningful overrepresentation of Republicans among X users. Our findings provide strong evidence of a partisan asymmetry in misinformation sharing which cannot be attributed to political bias on the part of raters, and indicate that Republicans will be sanctioned more than Democrats even if platforms transition from professional fact-checking to Community Notes.
An endosymbiotic origin of the crimson pigment from the lac insect
Symbioses with microorganisms expand the genetic and metabolic repertoire of many insects. The lac insect Kerria lacca (Hemiptera: Sternorrhyncha) is a phloem-feeding scale insect that is brightly colored due to the presence of natural polyhydroxy-anthraquinone pigments called laccaic acids. The deep red pigments possibly provide defense against pathogens and predators and are commercially important as dyes in textiles, lacquerware, and cosmetics. Laccaic acids are categorized as polyketides comprising an anthraquinone backbone decorated with tyrosine or its derivatives. However, the genetic basis of these pigments remains unknown, as insects are not known to produce aromatic polyketides or tyrosine de novo. Here, we sequence the genome of the lac insect and its two endosymbionts— Wolbachia and a hitherto unidentified, transovarially transmitted yeast-like symbiont (YLS). We found no evidence for the host or Wolbachia to be able to synthesize the pigments. The pigments and their precursors were also not detected in the host plant. Genomic, transcriptomic, and metabolomic analyses combined with fluorescence microscopy identified and characterized YLS as the sole producer of the pigment’s polyketide backbone and tyrosine moiety, demonstrating an endosymbiotic origin of the lac pigments. A nonreducing polyketide synthase gene cluster encoding the laccaic acid backbone was identified. Furthermore, the YLS genome encoded essential amino acids and vitamins that are deficient in the insect’s phloem diet. Experimental fungicide-treated insects exhibited reduced concentrations of laccaic acids and tyrosine, along with decreased body size and weight, indicating a mutualistic association between the lac insect and its YLS.