2026 Hankin Distinguished Lecture

Wednesday, November 11, 2026; 4 PM
Robb Hall, Hintz Family Alumni Center
Speaker: Dr. Todd Usher from

Hosted by: Jessica Wilson,  jrw412@psu.edu

Toward Robust and Efficient Reasoning with LLM Agents

Wednesday, November 18, 2026; 254 Health and Human Development Building
3:35-4:25 p.m.
Speaker: Aram Galstyan from

Abstract: Test-time scaling (TTS) aims to improve the reasoning capabilities of large language models (LLMs) and agents by allocating more compute at inference time. However, TTS introduces its own challenges, such as redundant reasoning, error propagation across multiple steps, and increased inference latency. In this talk, I will present our recent work addressing these challenges. First, I will show how selectively pruning redundant tokens from reasoning trajectories can improve accuracy without extra training. Next, I will introduce a framework for structured communication among parallel agents that detects and resolves factual inconsistencies during exploration while preserving diversity in reasoning trajectories. Finally, I will present an approach that uses idle time during tool execution for adaptive speculative planning, improving agent performance without increasing latency. Together, these results show how more effective use of test-time computation can make LLM agents more robust and efficient.

BIO: Aram Galstyan is a recognized leader in AI and machine learning with more than 25 years of experience and 150+ peer-reviewed publications in leading AI, machine learning, and natural language processing venues. He has a strong track record of leading interdisciplinary research teams, setting scientific strategy, and delivering large-scale, high-impact projects. His current research spans large language models, agentic systems, personalization, human-AI collaboration, and trustworthy AI.

As an Amazon Scholar in Amazon AGI Foundations, Aram provides technical leadership across initiatives in large language model safety, efficiency, and personalization. His work includes shaping research roadmaps, fostering collaboration between research and engineering teams, and developing academic partnerships that accelerate strategic AI initiatives.

At the USC Information Sciences Institute, Aram established and led the Machine Intelligence and Data Science group and directed the institute’s AI Division, its largest research unit. He led and mentored an organization of more than 100 researchers, engineers, faculty, postdoctoral scholars, and Ph.D. students. He also initiated and led major interdisciplinary research programs supported by DARPA, IARPA, NSF, and NIH, contributing to more than $60 million in sponsored research funding.

Hosted by: Lana Fulton,  lub18@psu.edu

Manipulating defect structure and behavior through segregation-induced complexion transitions

Wednesday, December 2, 2026; 3:35-4:25 pm
254 Health and Human Development
Speaker: Timothy Rupert from Johns Hopkins University

Abstract:

Defect engineering is known to be useful for manipulating material properties, yet often treats the crystalline flaws as temporary or metastable features. Recent advances have led to the identification of stable defect states or complexions, where local structure and chemistry can be controlled with thermodynamic variables such as composition or temperature. In this talk, we discuss how both grain boundaries and dislocations can host these complexions, with an eye toward increasing microstructural stability and tuning material response. First, the incorporation of amorphous intergranular films into a nanoscale grain structure is shown to solve many of the major, longstanding limitations of nanocrystalline metals. Notably, the ductility, toughness, strength, thermal stability, and radiation tolerance can all be simultaneously increased with the incorporation of amorphous grain boundary complexions, with examples shown for multiple alloy systems. Moreover, these materials can be fabricated into bulk forms with simple processing routes, demonstrating a clear path to commercial use. Next, complexion formation along dislocation lines offers a new pathway to manipulate higher-level, time-dependent mechanical properties associated with the collective behavior of dislocation populations and their interconnected networks. Local stress fields can modify the bulk solubility limits in nearby regions, leading to patterning of chemical order with nanoscale dimensions. Alternatively, local restructuring can occur to delocalize the dislocation core, resulting in new secondary defect structures. As a whole, this work lays the foundation for the engineering of defect structure and chemistry to design better nanomaterials, creating a new “defects-by-design” toolkit.

 

Bio:

Tim Rupert is a Professor of Materials Science and Engineering at Johns Hopkins University and the Director of the Hopkins Extreme Materials Institute (HEMI). Professor Rupert received a B.S./M.S. in Mechanical Engineering from Johns Hopkins University in 2007 and a Ph.D. in Materials Science and Engineering from MIT in 2011. His research focuses on uncovering new structure-property relationships in nanomaterials for structural and energy applications, as well as increasing the reliability and lifetime of nanostructured materials. The lab uses a combination of complementary computational and experimental techniques to tackle these challenges. Prof. Rupert has received the NSF CAREER Award, the DOE Early Career Award, the ARO Young Investigator Program Award, the Hellman Fellowship, the ASM International Bradley Stoughton Award for Young Teachers, the AIME Rossiter W. Raymond Memorial Award, the TMS Brimacombe Medal, and was elected as a Fellow of ASM International. He serves as an Editor for Acta Materialia and Scripta Materialia.

Hosted by: Lana Fulton,  lub18@psu.edu