3 papers
cs.LG2026
Linear Probes to Bayesian Manifolds: Geometry of Arithmetic in Large Language Models
Anshul Kumar, Neha Prasad, Manoj Kumar +2
Linear probes are a standard method to study geometric structures, such as helices and circles, from the activations of large language models (LLMs), and such structure is often ta…
cs.LG2025
AdaGradSelect: An adaptive gradient-guided layer selection method for efficient fine-tuning of SLMs
Anshul Kumar, Gagan Raj Gupta, Manisha Chawla
Large Language Models (LLMs) can perform many NLP tasks well, but fully fine-tuning them is expensive and requires a lot of memory. Parameter-Efficient Fine-Tuning (PEFT) methods s…
cs.SI2025
Flex-GAD : Flexible Graph Anomaly Detection
Apu Chakraborty, Anshul Kumar, Gagan Raj Gupta
Detecting anomalous nodes in attributed networks, where each node is associated with both structural connections and descriptive attributes, is essential for identifying fraud, mis…