Instituto de Física Interdisciplinar y Sistemas Complejos (IFISC)

15/10/2026

🧠 A new "thermostat" for virtual brains

Ever wondered how the human brain keeps its activity stable instead of crashing or going into overdrive?

An international study co-led by IFISC (UIB-CSIC) researcher Rubén Herzog, in collaboration with the Paris Brain Institute, introduces the HDMF model in Network Neuroscience.

Instead of forcing computers to manually balance neural activity, this new model adds an automatic biological "thermostat", a homeostatic plasticity rule, allowing the virtual brain to adjust itself in real time.

🔗 https://ifisc.uib-csic.es/en/news/digital-twins-of-the-human-brain/

15/10/2026

🛤️ Torna Ciència a Tot Tren!

🗣️ Gaudeix d'una experiència única a bord del Tren De Soller, amb minixerrades de científics de diverses institucions públiques de de les Illes Balears.

🌳🍊 Completa la jornada amb una passejada pel centre del poble i visita de franc el MUCBO Jardí Botànic de Sóller-Museu Balear de Ciències Naturals.

📢 Novetat! Les inscripcions s'obrin avui, 14 d'octubre, i hi haurà un termini de 48 hores per apuntar-s'hi. Enguany, les places s'assignaran per sorteig.

➡️ Una activitat organitzada per la i la Delegación del CSIC en las Illes Balears, amb el finançament de la F E C Y T · Ciencia e Innovación ·.

Ministerio de Ciencia e Innovación Institut Mediterrani d’Estudis Avançats Instituto de Física Interdisciplinar y Sistemas Complejos (IFISC) Idisba Ajuntament de Sóller Instituto Geológico y Minero de España

🔗 Informa-te'n i participa-hi! https://culturacientifica.uib.cat/projectes/curs-2026-2027/ciencia-a-tot-tren-2026

24/09/2026

| 💬 Investigadors de l'Instituto de Física Interdisciplinar y Sistemas Complejos (IFISC) proposen una nova manera d'estudiar com canvien les opinions en la societat.

➡️ L'estudi adopta un enfocament nou que incorpora eines modernes de la termodinàmica estocàstica de la informació per descriure el comportament social humà.

🔗 Llegeix la notícia! https://diari.uib.cat/w/com-les-lleis-termodinamiques-poden-descriure-el-comportament-social-huma?title_redirect=Portada&redirect=%2F

24/09/2026

Can the principles of physics help us understand how public opinion shifts? 🧠💬

A new study by IFISC (UIB-CSIC) researchers Luis Irisarri, Lucas Trigal, Raul Toral and Gonzalo Manzano, published in Nature Communications, applies stochastic thermodynamics to social dynamics. They prove that opinion flows follow a "second-law" rule without needing concepts like energy or temperature.

The team shows how the tug-of-war between fitting in (conformity) and standing out (anticonformity) can trigger abrupt social tipping points, much like phase transitions in physics, explaining how societies suddenly shift from polarization to consensus. This framework offers a fresh, data-driven way to study collective behavior.

🔗 https://ifisc.uib-csic.es/en/news/how-thermodynamical-laws-can-describe-human-social-behavior/

07/09/2026

Calling all researchers in computational neuroscience and complex systems.

IFISC (UIB-CSIC) and the Department of Psychology at the Universitat de les Illes Balears - UIB are thrilled to host the Workshop on Whole-Brain Models and High-Order Interactions for Brain Health.

📅 When: November 17–20, 2026
📍 Where: Complex Balear de Recerca (Parc BIT), Palma de Mallorca, Spain

An exclusive, hands-on event bringing together 50 selected researchers (Master's students, PhDs, and postdocs) to tackle modern neuroscience's central challenges using Python workflows, machine learning, and information-theoretic methods. Featuring world-class international speakers, flash talks, poster sessions, and a collaborative project sprint.

🔹 Key Details:

- No registration fee (limited spots available!)
- 4-day intensive, in-person program
- Focus on whole-brain modeling, neuroimaging (fMRI, M/EEG), and collective brain dynamics.

⏳ Application Deadline: September 15, 2026.

👉 Check out all the details and submit your application: https://rubenherzog.github.io/workshop-wbm-hoi-2026/

10/08/2026

🏨 How does the tourism market react to an exceptional event like the 2026 solar eclipse?

A new undergraduate thesis at Universitat de les Illes Balears, supervised by Pere Coelt (IFISC, UIB-CSIC), applies data analytics and statistical physics to examine accommodation trends. Findings reveal a stark contrast: while peninsular regions see massive price hikes and depleted availability, the Balearic Islands' peak season capacity acts as a natural buffer, keeping the market stable.

🔗 Full press release: https://ifisc.uib-csic.es/en/news/study-reveals-why-the-2026-eclipse-had-a-muted-impact-on-balearic-accommodations/

26/06/2026

📣 Time is running out! Preregistration is now open for the Master’s in Physics of Complex Systems at Universitat de les Illes Balears.

Explore cutting-edge research, connect with an international network of experts, and build the skills to tackle today’s most challenging systems, from brain dynamics to social networks.

📆 Apply by June 30: https://www.uib.eu/Learn/estudis-de-master/master/MFS3/

▶ Mobility fellowships available!: https://ifisc.uib-csic.es/en/master/fellowships-for-the-masters-degree/

Fellowship deadline: July 12

16/06/2026

Do we really need complex models to understand how groups interact? 👥🕸️

In network science, there's been a big shift toward using hypergraphs to map things like peer pressure and collective decision-making. But it's sparked a good debate: when do we actually need that extra mathematical complexity, and when are simpler networks enough? A new study published in Physical Review Research by the team at IFISC (UIB-CSIC) shares some insights on this.

The researchers (Jaume Llabrés Rubio, Raul Toral, Maxi San Miguel and Federico Vázquez) showed that many social impact models on hypergraphs can actually be simplified into straightforward, pairwise interactions without losing any microscopic accuracy. This means we can often use classic, well-established network tools to study complex group dynamics.

🔬 Read the full breakdown: https://ifisc.uib-csic.es/en/news/do-group-interactions-really-require-higher-order-models/

09/06/2026

🏖 Yesterday the IFISC (UIB-CSIC) staff celebrated our usual Summer Solstice at Son Serra de Marina. Beach, sport, good food and interesting conversations!

☀ Happy Summer Solstice!

25/05/2026

🚨 What if a little chaos could help artificial intelligence learn faster?

A new study from IFISC (CSIC-UIB), published in Physical Review Research, shows that neural networks can train more efficiently when operating near the so-called “edge of chaos”: a regime between order and disorder where exploration and optimization are better balanced.

Far from being harmful, a controlled amount of chaotic behavior may actually speed up AI learning.

The study was carried out by Pedro Jiménez González, Miguel Cornelles Soriano, and Lucas Lacasa.

Read more 👇
https://ifisc.uib-csic.es/en/news/a-touch-of-chaos-can-help-artificial-intelligence-learn-faster/

🖥️ El caos controlado acelera el aprendizaje de la

⌨️Las redes neuronales artificiales son sistemas computacionales clave para la inspirados en las neuronas biológicas

🔬 Un estudio del Instituto de Física Interdisciplinar y Sistemas Complejos (IFISC) muestra que procesan la información de manera más eficiente cuando operan al inicio del comportamiento caótico

🔎Les permite refinar las soluciones conocidas y explorar nuevas posibilidades entre el vasto número de configuraciones posibles

👉 http://tiny.cc/7a04101