bsky.app/profile/grf....
29.01.2026 17:55 β π 0 π 0 π¬ 0 π 0@grf.bsky.social
Research Institute on Applied Mathematics and Systems at the National Autonomous University of Mexico. Currently thinking about collective behavior and cognition. https://orcid.org/0000-0001-7175-3905
bsky.app/profile/grf....
29.01.2026 17:55 β π 0 π 0 π¬ 0 π 0A pleasure to chat with Matt Galloway in The Current, earlier this morniing, about our recently published paper on spider monkeys' collective intelligence www.cbc.ca/listen/live-...
29.01.2026 17:53 β π 1 π 0 π¬ 1 π 0I'm really happy to share this paper showing a relationship between social integration and offspring survival in female spider monkeys. I was lucky to work on this with Cristina Jasso, @grf.bsky.social and a great group of collaborators.
link.springer.com/article/10.1...
Spider monkeys use collective intelligence to find food: www.earth.com/news/spider-...
27.01.2026 16:33 β π 2 π 1 π¬ 0 π 0Impressive research and impressive spider monkeys!
#primates #cognition #information #network #sharing
https://www.theguardian.com/science/2026/jan/25/spider-monkeys-found-to-share-insider-knowledge-to-help-locate-best-food?utm_source=dlvr.it&utm_medium=bluesky&CMP=bsky_gu
Spider monkeys crowdsource best places to eat in forest www.hw.ac.uk/news/2026/sp...
26.01.2026 16:40 β π 1 π 0 π¬ 0 π 0npj Complex.: Uncovering complementary information sharing in spider monkey collective foraging using higher-order spatial networks
https://www.nature.com/articles/s44260-025-00060-0
ππ§ New paper in npj Complexity: complementary information sharing in fission-fusion dynamics.
Video explainer: youtu.be/PIAhcLWqsO8?...
Full paper (open access): doi.org/10.1038/s442...
Higher-order spatial networks enable distributed foraging knowledge in heterogeneous environments.
π
Fission-fusion dynamics implies distributed information processing analogous to that occurring in "liquid brains."
Collective intelligence through complementary knowledge sharing! π§ β¨
Collaborators:Ross S. Walker, @mattjsilk.bsky.social, Denis Boyer and @sandrateles-esmag.bsky.social
Maximal simplicial degree centrality shows sigmoidal scaling with simplex size: intermediate-sized simplices (4-6 individuals) are most central to the network.
Centrality is independent of age, sex, or immigration status β knowledge asymmetry driven by mobility, not individual traits.
These holes persist across multiple filtration values, indicating robust complementarity.
Even when including less redundant interactions, spatial overlaps don't form trivial structures - unique knowledge pockets remain across various subset sizes.
Persistence barcodes for the simplicial complexes in each season, including the simplices of individual core ranges formed by decreasingly redundant interactions.
We constructed filtered simplicial complexes using Ξ± as a filtration parameter (scaled by deviation from w*).
Persistent homology reveals topological holes, which we interpret as structural signatures of complementary information distribution πΈοΈ
Scaling relationship between the number of individual core ranges and its intersection/union ratio, w. Each dot corresponds to a particular set of a given number of individuals. The solid black line corresponds to the scaling relationship.
We found that in some cases, the observed w values closely follow the predicted convex decreasing function of set size n.
Smaller sets (2-6 individuals) show higher variance, suggesting this range is most dynamic for knowledge complementarity.
Conceptual illustration of the degree of redundancy and uniqueness shown by four hypothetical overlapping individual core ranges.
We derived an optimal intersection/union ratio (w*) that maximizes information transfer between n individuals:
w* = 1/(n+1)
This balances redundant overlap (coincidence opportunities) with unique knowledge (areas known exclusively by subsets).
Two sets of overlapping core ranges showing both redundant and unique areas
We estimated individual core ranges (60% utilization distributions) from 6 years of location data in Yucatan, Mexico.
We assumed that these core ranges represent an individual's spatial knowledge for a given season, while partial overlaps create opportunities for information transfer.
Spider monkeys (Ateles geoffroyi) exhibit high fission-fusion dynamics - individuals form subgroups that join and split frequently.
We asked whether distributed spatial knowledge across individuals enabled collective processing of foraging information beyond individual capacity.
ππ§ New paper in npj Complexity: complementary information sharing in fission-fusion dynamics.
Video explainer: youtu.be/PIAhcLWqsO8?...
Full paper (open access): doi.org/10.1038/s442...
Higher-order spatial networks enable distributed foraging knowledge in heterogeneous environments.
π
Thrilled to see this paper out, two years after starting our collaboration at @divintelligence.bsky.social
30.12.2025 19:58 β π 9 π 3 π¬ 0 π 0Jane Goodall with the rest of participants at the Animal Social Complexity and Intelligence conference, Chicago, 2000
Always uniquely funny and inspiring. Here at one of the first conferences I attended as a grad student, the Animal Social Complexity and Intelligence conference in Chicago, in 2000. Thank you Jane Goodall.
01.10.2025 19:04 β π 2 π 0 π¬ 0 π 0bsky.app/profile/grf....
04.06.2025 18:14 β π 0 π 0 π¬ 0 π 0New paper in @behavecol.bsky.social (link: academic.oup.com/beheco/advan...)
led by PhD student @marcofele.bsky.social using hard-won data from our amazing baboon team. Our @swanseauni.bsky.social press release:
www.swansea.ac.uk/press-office...
We introduce the idea of a "social spandrel".....
Uncovering complementary information sharing in spider monkey collective foraging using higher-order spatial networks arxiv.org/abs/2505.01167
28.05.2025 07:25 β π 4 π 2 π¬ 1 π 05/5
combined knowledge has complementarity: each individual fills in parts of the puzzle that others miss. We argue that the ever-changing structure of these groups helps them collectively adapt and make the most of their shared knowledge in unpredictable environments.
π π§© π = π§
4/5
to share key information, but also enough unique knowledge to cover more ground. To model these interactions, we used a mathematical approach (simplicial complexes) that captures how multiple animals exchange information at once. This revealed 'gaps' in different dimensionsβmeaning the groupβsβ¦
3/5
track changes in their environment better than any individual could on its own. We studied how much individual animals' core ranges overlap, treating these overlaps as a way to measure shared vs. unique knowledge of food sources. Some range combinations strike a balance: they have enough overlap
2/5
Groups can often process information together more effectively than any single member could alone. In species with fission-fusion dynamicsβwhere animals frequently split up and reunite (like dolphins or chimpanzees)βsharing different bits of knowledge about food locations helps the whole groupβ¦
π§ͺNew preprint: βUncovering complementary information sharing in spider monkey collective foraging using higher-order spatial networksβ in collaboration with Ross Walker, @mattjsilk.bsky.social, Denis Boyer and @sandrateles-esmag.bsky.social
arxiv.org/abs/2505.01167
π§΅ 1/5
Preprint: The causal role of synergy in collective problem solving. Great collaboration with DISI fellows @ketikagarg.bsky.social @culturologies.co, Zara Anwarzai and Hannah Dromiack
23.04.2025 18:40 β π 4 π 2 π¬ 0 π 0With colleagues from our social complexity seminar, we joined the debate in BBS on "What is a society" by M. Moffett. We argue that a society can emerge from social interaction patterns without the need for establishing an a priori limit on who actually belongs to it doi.org/10.1017/S014...
09.04.2025 10:35 β π 1 π 1 π¬ 0 π 0βEverything is political, especially the things that people tell you are not political. Those are the most political of allβ
05.03.2025 10:02 β π 3 π 0 π¬ 0 π 0