• Geometric particle swarm optimization in Program Trace Optimization 

  Moraglio, A; McDermott, J (Springer Nature, 2025)

  On the 30th anniversary of Particle Swarm Optimization (PSO), we present a novel integration of Geometric PSO (GPSO) into the Program Trace Optimization (PTO) framework. GPSO extends PSO to diverse representations in a ...

• Composable Interventions for Language Models 

  Kolbeinsson, A; O'Brien, K; Huang, T; et al. (International Conference on Learning Representations, 2025)

  Test-time interventions for language models can enhance factual accuracy, mitigate harmful outputs, and improve model efficiency without costly retraining. But despite a flood of new methods, different types of interventions ...

• Exploring Sentinel-1 Radar Polarisation and Landsat Series Data to Detect Forest Disturbance from Dust Events: A Case Study of the Paphos Forest in Cyprus 

  Theocharidis, C; Eliades, M; Kolokoussis, P; et al. (MDPI, 28 February 2025)

  Monitoring forest health has become essential due to increasing pressures caused by climate change and dust events, particularly in semi-arid regions. This study investigates the impact of dust events on forest vegetation ...

• SPAM: Spike-Aware Adam with Momentum Reset for Stable LLM Training 

  Huang, T; Zhu, Z; Jin, G; et al. (International Conference on Learning Representations, 2025)

  Large Language Models (LLMs) have demonstrated exceptional performance across diverse tasks, yet their training remains highly resource intensive and susceptible to critical challenges such as training instability. A ...

• Enhancing Robust Fairness via Confusional Spectral Regularization 

  Jin, G; Wu, S; Liu, J; et al. (International Conference on Learning Representations, 2025)

  Recent research has highlighted a critical issue known as robust fairness", where robust accuracy varies significantly across different classes, undermining the reliability of deep neural networks (DNNs). A common approach ...

View more