The team’s work pivots on a deceptively simple behavioral dichotomy observed in this species: the switch between “explore,” characterized by swift, erratic movements to gather sensory data, and “exploit,” slower, focused behaviors aimed at executing known tasks. These alternating modes are not only critical for survival in complex, often unpredictable environments but may also provide clues to designing adaptive robotics and uncovering hidden layers of brain function. The fish owe their ethereal ghost-like movements to their electrosensory system, which they employ to navigate and hunt in murky waters, generating weak electric fields to sense obstacles and prey alike.

Previous foundational work, published in 2023 in the prestigious journal Nature Machine Intelligence, identified remarkably consistent patterns of explore/exploit behavior not just in glass knifefish but across a broad phylogenetic spectrum stretching from unicellular amoebas to humans. This cross-species similarity suggests that the neural principles governing such decision-making are deeply conserved, pointing to fundamental biological rules. In those studies, researchers meticulously analyzed velocity profiles and behavioral modes, identifying bimodal patterns reflective of alternating exploratory bursts and exploitative glide phases.

What sets the new project apart is its expanded temporal scope and technical sophistication. Earlier experiments captured brief snapshots—approximately 40 seconds per trial—insufficient to tease apart the nuanced dynamics underlying mode switches. Now, with trials extended to 10 minutes, the team is poised to observe subtler behavioral metrics, such as the length of movement bursts, spatial correlations within the constrained environment, and potentially the triggers that precipitate shifts between modes. These detailed observations will enable a granular understanding of behavior unfolding over time scales that better simulate the challenges animals face in the wild.

Central to advancing these insights is the incorporation of cutting-edge neural recording techniques. For the first time, electrodes implanted in the brains of glass knifefish will capture real-time neural activity concurrent with behavior. This endeavor, led by biologist Eric Fortune at NJIT, overcomes a formidable obstacle faced by earlier studies where neural correlates were inferred indirectly. Such direct neural-behavioral mapping empowers the team to investigate the mechanistic substrate of decision making—probing how fluctuating internal states and sensory uncertainty interact within neural circuits to influence mode switching.

The hypothesis under scrutiny posits that the fish’s decision to switch modes hinges on an internal estimation of uncertainty. If the fish perceives ambiguity about its location within the tube, it triggers an exploratory movement burst to actively gather sensory data, reducing uncertainty before resuming goal-directed behavior. This theory aligns with broader computational neuroscience perspectives that conceptualize the brain as an inference engine constantly balancing information acquisition against action execution. To model these complex processes, the team integrates sophisticated machine learning algorithms developed at the University of Minnesota, transforming behavioral and sensory datasets into mathematical functions capable of capturing underlying causal relationships.

Kathleen Hoffman, professor of mathematics and statistics at UMBC and a pivotal figure in the project, underscores the importance of combining human intuition with formal computational tools. Her approach begins with manual pattern recognition—visually parsing velocity and position data to hypothesize behavioral motifs—before encoding these observations into automated analyses to rigorously test their recurrence across the expanded dataset. This iterative interplay of qualitative and quantitative techniques exemplifies how mathematics and statistics can serve as a bridge between raw biological data and interpretable models.

Noah Cowan highlights the collaborative synergy driving this research forward. His laboratory’s decade-long struggle to decipher glass knifefish behavior finds new momentum through the assembly of this multidisciplinary “dream team” of neuroscientists, engineers, mathematicians, and computer scientists. Each expert contributes a vital piece of the puzzle, from behavioral assays and neural physiology to data analytics and algorithm development. The project’s holistic scope exemplifies modern neuroscience’s embrace of integrative methods to unravel complex brain-body-environment interactions.

Beyond its scientific ambition, the project holds transformative potential for technological innovation. In robotics, mimicking the intermittent sensing tactics observed in these animals could revolutionize navigation in uncertain or hazardous terrains—such as disaster zones—where continuous scanning is costly and inefficient. By understanding the decision rules underlying when to explore versus exploit, engineers can design robots that adapt their sensor usage dynamically, conserving energy and optimizing task performance. Moreover, the discovery of common neural strategies across species hints at broader biomedical applications, including new insights into neurological conditions characterized by impaired decision-making, though such clinical implications remain speculative at this stage.

An equally important facet of the project is its educational mandate. By integrating undergraduate students into the data visualization and analysis pipeline, the team fosters hands-on interdisciplinary training, preparing the next generation of researchers to tackle complex, multifaceted questions at the intersection of biology, mathematics, and engineering. This educational dimension ensures that the ripple effects of the research will extend beyond immediate scientific outcomes, nurturing a culture of collaborative, cross-domain inquiry.

In sum, this initiative encapsulates a profound scientific endeavor: to decode the algorithms etched into neural circuits that orchestrate the fundamental trade-off between exploring new information and exploiting known resources. The weakly electric glass knifefish, with its elegant dichotomy of movement modes and accessible neural architecture, offers a compelling window into these processes. As data collection scales up and analytical methods mature, the team anticipates breakthroughs that could not only complete a decade-long quest but also catalyze advances in robotics, neuroscience, and our fundamental understanding of decision-making.

The cutting-edge fusion of experimental neuroscience, theoretical mathematics, and machine learning in this project highlights the evolving nature of scientific discovery—one that transcends disciplinary silos to tackle the complexity of living systems. With each carefully recorded flicker of the glass knifefish’s electric field, scientists edge closer to deciphering the language of the brain’s internal compass, illuminating pathways that govern behavior from the simplest organisms to ourselves.

Subject of Research: Neural mechanisms underlying explore/exploit behavior in animals, studied through the weakly electric glass knifefish as a model.

Article Title: [not explicitly stated in the source]

News Publication Date: [not explicitly stated in the source]

Web References:

• Collaborative Research in Computational Neuroscience (CRCNS): https://www.nsf.gov/funding/opportunities/crcns-collaborative-research-computational-neuroscience
• Prior related publication in Nature Machine Intelligence: https://www.nature.com/articles/s42256-023-00745-y
• UMBC story on animal decision-making: https://umbc.edu/stories/animal-decision-making-with-robotics-applications/

References:

• Hoffman, K., Cowan, N., et al. (2023). Explore/Exploit behavior across species. Nature Machine Intelligence.
  (Additional references not specified)

Image Credits: Noah Cowan

Keywords:
Explore/exploit decision-making, glass knifefish, neuroscience, computational neuroscience, machine learning, behavioral neuroscience, neural recordings, robotics, sensory uncertainty, interdisciplinary research

Assistant Professor Amelie Schiprowski’s research dives into the complexities of labor economics during a time of demographic transformation that is reshaping global labor markets. With aging populations and shifting workforce compositions, companies face increasing challenges in recruiting young talent despite persistently high youth employment rates. Schiprowski’s ERC-funded project, “Entry-Level Hiring in Tightening Labor Markets: Frictions, Firm Heterogeneity and Public Policy” (ENTRYHIRE), seeks to unravel the intricate matching processes between emerging workers and employers under conditions of labor market stress. Utilizing comprehensive empirical data from the German apprenticeship system, her research involves sophisticated econometric models designed to elucidate how firms adapt their hiring and training practices in response to constrained labor supply. Through this, Schiprowski aims to produce evidence-based policy recommendations that can improve labor market efficiency and support smoother workforce entry for young workers.

Meanwhile, evolutionary biologist Dr. Moritz Lürig embarks on a pioneering journey to decode the vibrant and complex wing coloration patterns exhibited by butterflies and moths—the Lepidoptera—through an innovative blend of artificial intelligence and evolutionary biology. His project, “The Evolution of Wing Coloration in Lepidoptera” (EWINCOL), leverages millions of digitized images sourced from natural history museums worldwide to construct a comprehensive digital atlas of wing color patterns. This “wing atlas” represents a revolutionary tool to explore the evolutionary trajectories of color and shape diversity in these insects. Using deep learning algorithms, Lürig seeks to identify correlations between coloration changes and speciation events, understand whether and how wing colouration and morphology have co-evolved or diverged, and investigate the influence of environmental variables such as habitat light conditions, temperature gradients, and geographic distribution on phenotypic diversity. This approach promises novel insights into the genetic and ecological drivers of evolutionary biodiversity.

In the realm of computer science, Professor Lucie Flek tackles one of the most urgent challenges in artificial intelligence: imbuing AI systems with genuine social intelligence and empathy. Her ERC-funded project, titled LLMpathy, aims to transcend the superficial mimicry of human emotions currently prevalent in AI conversational agents by equipping language models with the capability to structure, justify, and explain human thoughts and feelings in a causally coherent manner. By integrating advanced machine learning techniques with longitudinal psychological research, Flek’s project will develop personalized AI profiles linking traits, values, emotions, and behaviors to enable models to reason more transparently and respond with nuanced social understanding. The project also involves creating simulated environments where AI agents engage in complex social interactions such as conflict resolution and negotiation, setting the stage for empirical assessment and refinement of perspective-taking and goal-oriented behaviors in AI. This work is not only foundational for enhancing human-AI interaction but also essential for ensuring ethical AI development aligned with transparency, trustworthiness, and the forthcoming EU AI regulatory framework.

The University of Bonn’s interdisciplinary ethos is highlighted by the diversity of these ERC projects spanning labor economics, evolutionary biology, and artificial intelligence, each addressing pressing scientific questions with innovative methodologies. Schiprowski’s empirical investigation into labor market mechanisms offers vital insights into socioeconomic policies amid demographic shifts, promising to inform effective public interventions that safeguard youth employment and a sustainable workforce. Lürig’s data-driven evolutionary analysis leverages cutting-edge AI to unlock centuries-old genetic mysteries encoded in butterfly and moth wing patterns, potentially transforming our understanding of natural selection and adaptation. Flek’s socially conscious AI research confronts the limitations of current language models, pioneering avenues toward machines that can genuinely empathize and reason about human emotions, crucial for applications spanning healthcare, education, and digital communication.

The ERC Starting Grant, universally recognized as one of the most competitive funding programs globally, empowers early-career scientists with a proven scientific track record to pursue visionary ideas over five years, with up to €1.5 million in grant support. Applicants from any nationality are eligible, provided they demonstrate excellence in their fields and submit rigorous proposals via their host institutions. This grant not only fuels scientific discovery but also fosters the development of the next generation of leading researchers prepared to tackle complex global challenges.

Amid ongoing global discussions on labor market resilience, biodiversity conservation, and the ethical deployment of AI, the work undertaken by these three scholars at the University of Bonn stands out for its compelling blend of technical rigor and societal relevance. Schiprowski’s nuanced analyses of labor market frictions will aid policymakers in adapting to the realities of shrinking youth labor pools and evolving company hiring behaviors. Lürig’s methodological fusion of AI and museum collections exemplifies the transformative power of technology in natural sciences, enabling large-scale comparative analyses previously unattainable. Meanwhile, Flek’s exploration of personalized, ethically transparent AI promises to revolutionize human-machine interaction, addressing concerns about emotional manipulation and bias while leveraging AI’s potential as a socially competent assistant.

The synergy of expertise at the University of Bonn further amplifies the impact of these projects. For instance, Flek’s work blends computer science with psychology and ethics, reflecting a transdisciplinary approach vital for developing trustworthy AI. Schiprowski’s integration with the “ECONtribute” Cluster of Excellence couples empirical macroeconomic research with policy dialogues, while Lürig’s collaboration with international museums showcases the global scope of biodiversity studies empowered by AI. This collaborative environment nurtures cross-pollination of ideas and facilitates both theoretical advancements and practical applications.

As these projects commence, the scientific community and broader public alike can look forward to transformative findings that advance knowledge across multiple domains. Schiprowski’s labor market insights are expected to inform economic models that better account for demographic trends and firm heterogeneity, thus improving labor policy design. Lürig’s digital wing atlas will serve as a foundational resource for future evolutionary studies and conservation efforts by providing a detailed morphological and ecological database. Concurrently, Flek’s LLMpathy initiative anticipates breakthroughs in AI architecture that marry cognitive and affective computing, fostering machines capable of richer human-like social engagement.

Ultimately, the achievements of these researchers underscore a larger paradigm within contemporary science: that interdisciplinarity, data-driven methods, and socially embedded research are critical to tackling the multifaceted challenges of the 21st century. The University of Bonn’s success in securing multiple ERC Starting Grants not only honors individual academic excellence but also reflects the institution’s vibrant research culture poised to make enduring contributions to science, technology, and society.

For more detailed inquiries and ongoing updates, the media contacts associated with each project remain accessible. Assistant Professor Amelie Schiprowski can be reached through the ECONtribute Cluster of Excellence, Professor Lucie Flek is affiliated with the Lamarr Institute and Bonn-Aachen International Center for IT (b-it), while Dr. Moritz Lürig is connected with the Florida Museum of Natural History. The continued dissemination of their work promises to inspire interdisciplinary innovation and the responsible advancement of knowledge at the crossroads of economics, biology, and artificial intelligence.

Subject of Research: Labor economics focusing on labor market entry mechanisms; evolutionary biology investigating wing coloration evolution in Lepidoptera; computer science research on socially intelligent AI and language models.

Article Title: University of Bonn Researchers Awarded ERC Starting Grants to Advance Labor Economics, Evolutionary Biology, and Socially Aware Artificial Intelligence

News Publication Date: (Not provided)

Web References: (Not provided)

References: (Not provided)

Image Credits: Collage by Max Waidhas/University of Bonn; Marc Thürach/ECONtribute; Kristen Grace/Florida Museum of Natural History

Keywords: ERC Starting Grant, labor economics, demographic change, apprenticeship market, evolutionary biology, butterfly wing coloration, Lepidoptera, artificial intelligence, socially intelligent AI, language models, AI ethics, interdisciplinary research

The InTheGra grant offers a unique funding opportunity for up to 24 months, providing a vital, non-renewable scholarship intended to catalyze cutting-edge research projects. Upon completion of this fellowship, recipients are strongly encouraged to secure permanent or visiting positions either within São Paulo’s host institutions or internationally, thereby fostering sustained academic engagement and knowledge exchange. The program’s structure underscores FAPESP’s commitment to bridging the gap between international excellence and São Paulo’s dynamic scientific community.

To be eligible, candidates must submit a meticulously developed research project aligned with FAPESP’s Thematic Projects framework, which demands bold, interdisciplinary, and high-impact research proposals. These projects must be led by a Principal Investigator (PI) with a PhD or equivalent qualification obtained at least ten years prior and have demonstrable success working within internationally competitive research groups outside Brazil. This ensures that funded projects are driven by seasoned experts who bring a wealth of experience and a proven track record of scientific leadership.

The design of the InTheGra program inherently promotes the involvement of multidisciplinary teams comprising co-principal and associate researchers as well as postdoctoral fellows, graduate students, and technical staff. Such comprehensive research groups are poised to push the boundaries of knowledge, addressing complex scientific questions that require diverse expertise and collaborative synergy. This approach aligns with international trends emphasizing integrative and cross-disciplinary research tackling global challenges.

An essential element of the InTheGra initiative is the embedded support structure expected from host institutions within São Paulo. Selected candidates’ affiliations must guarantee full administrative backing, unrestricted access to advanced research facilities, and integration of the grant holder into the faculty framework of the host institution’s graduate programs. These provisions collectively ensure that researchers receive a robust environment conducive to innovative scientific inquiry and academic mentoring.

The timeline for the current call is strategically phased to maximize the quality of submissions and candidate selection. Pre-proposals must be submitted by June 30, after which a shortlist of pre-qualified candidates will be announced in late July. These selected candidates will subsequently be invited to submit full proposals by the September 28 deadline. Importantly, all applications and submissions will be managed through FAPESP’s SAGe platform, demonstrating the foundation’s commitment to a streamlined and transparent evaluation process.

What sets the InTheGra program apart from conventional fellowship schemes is the emphasis on institutional integration and the post-fellowship career trajectory of awardees. FAPESP encourages grantees to consolidate their careers either by securing tenure-track positions internationally or by returning to São Paulo, reinforcing the exchange of ideas and expertise between global and local scientific communities. This pathway illustrates a forward-thinking approach to science funding, blending mobility with institutional attachment.

From a technical standpoint, the program aligns with current efforts worldwide to internationalize research by attracting diaspora scientists and foreign experts into local ecosystems. This internationalization not only raises the competitive standard of research output in São Paulo but also fosters collaborations with high-impact global centers of knowledge. This is particularly relevant in the age of complex scientific challenges that demand mobilization of global expertise, advanced methodologies, and innovative technologies.

The program’s call for “thematic” projects further underscores a focus on ambitious, big-picture research that transcends narrow disciplinary boundaries. It expects projects to be capable of producing scientific breakthroughs or transformative knowledge that can influence policies, technologies, or society at large. This characteristic positions the InTheGra grants as a potentially catalytic force within the wider Brazilian and international funding landscapes.

Moreover, the multidisciplinary nature of proposed projects emphasizes the involvement of a spectrum of contributors—from senior scientists to early-career researchers. This layered research team architecture drives capacity-building within São Paulo’s academic environment, ensuring the nurturing of future generations of scientists equipped with international outlooks and robust methodological skills.

The emphasis on securing institutional employment post-fellowship highlights a sustainability model keyed to long-term academic and scientific integration. By encouraging tenure options or visiting researcher positions, the program seeks to maintain the momentum created during the funded period, avoiding the typical disconnect many international scholars face when reintegrating into local research environments.

In summary, FAPESP’s InTheGra initiative exemplifies a strategic, scientifically rigorous, and internationally aware funding mechanism designed to propel São Paulo to the forefront of global research innovation. This program not only provides necessary financial resources but also integrates researchers into thriving academic communities and sets a precedent for the kind of scientific excellence and collaboration essential in today’s interconnected world. With carefully structured application processes, clear eligibility criteria, and a focus on high-impact outcomes, the InTheGra call invites researchers to embark on ambitious projects that can redefine scientific frontiers.

Institutions and researchers interested in this transformative opportunity are encouraged to review the detailed program parameters and submission guidelines available on FAPESP’s official portal. The call stands as a beacon of excellence and a call to global scientific talent to contribute to the vibrant research tapestry of São Paulo, shaping the future of science regionally and globally.

Subject of Research: Research funding and international research collaboration focusing on high-impact thematic projects with international Principal Investigators.

Web References:
https://fapesp.br/17523

Keywords:
Research funding, Academic researchers, Human resources, Educational institutions, Science policy, Research management, International scientific collaboration