Dr. Blaž Bertalanič and Dr. Carolina Fortuna of the Department of Communication Systems were awarded an Industry Spotlight at the AI Engineer World's Fair conference and invited for presentation at the AI Engineer World's Fair in San Francisco with more than 6,000 developers from every frontier AI lab. Their finding cuts against how multi-agent systems are being built today. Multi-agent LLM systems, i.e. teams of AI agents that debate and vote on answers, are widely deployed on the assumption that more agents means more capability. Dr. Bertalanič and Dr. Fortuna have shown quantitatively that this assumption often fails: agent teams show sharply diminishing returns as they scale. Their study found that unguided debate among identical agents frequently underperforms simple isolated self-correction, burning 2–3 times more computation for equal or lower accuracy—undermined by agents conforming to the majority, abandoning correct reasoning under peer pressure, and by voting that discards correct answers already generated. The finding points to architectural diversity, not sheer agent count, as the key lever for scaling multi-agent AI systems in production. |
On July 15, 2026, the new U.S. Ambassador to Slovenia, Her Excellency Asel K. Roberts, visited the Jožef Stefan Institute. The director of the Jožef Stefan Institute, Prof. Dr. Leon Cizelj, began by introducing her to Slovenia’s largest scientific and research institution, which undoubtedly offers excellent opportunities for cooperation with the United States in various fields. “The fields of artificial intelligence, the TRIGA reactor and nuclear technologies, quantum technologies and superconductivity, as well as projects related to the CERN and ITER accelerators, are at the forefront of the Jožef Stefan Institute’s work. We are already collaborating with the U.S., and during today’s visit, we explored possibilities for even better expanding this cooperation and collaboration with Ambassador Roberts,” said the Institute’s director, Prof. Dr. Leon Cizelj. Most of the JSI’s formal international ties are linked to the European Research Area; cooperation with the U.S. takes place primarily through joint research projects, scientific publications, researcher exchanges, and participation in major international research infrastructures. Both Cizelj and Roberts shared their hope for continued broad cooperation. |
Disordered superconducting materials are widely used in quantum devices—qubits, microwave resonators, photon detectors, and more. Their performance is limited by unexplained energy loss at low temperatures. Prof. Mikhail Feigelman of the Department of Complex Matters, together with his colleague Anton V. Khvalyuk, has developed a new microscopic theory that explains how energy loss depends on temperature (T) and frequency (ω) under typical operating conditions, when both T and ω are low compared to superconducting energy scales. The key discovery, published in the journal Physical Review Letters, is that losses are dominated by localized collective modes arising from irregularities in the superconducting state—another manifestation of disorder. The energy loss to these modes increases rapidly with frequencies and decreases with temperature—an unusual consequence of the local structure of these modes. The theory explains recent experiments on thin films of materials such as indium oxide, titanium nitride, and niobium nitride, and points to practical strategies to reduce energy loss in quantum hardware. |
|
