2 papers
cs.LG2026
MOSAIC: Addressing Disproportionate Task-Modality Dependencies via Modality-Partitioned Gradient Routing
Aryan Chowdhury, Yizheng Sun, Thomas Carroll +1
In multi-task settings, combining multiple modalities allows task representations to draw on signals that a single modality alone would not capture. However, individual tasks may d…
cs.CL2026
Fewer Heads, Fewer Memories: Pruning Redundant Attention Heads Suppresses Verbatim Memorisation
Aryan Chowdhury, Riza Batista-Navarro
Large language models have been shown to memorise and reproduce verbatim sequences from their training data, raising concerns about privacy leakage, intellectual property exposure,…