Search papers, labs, and topics across Lattice.
Singapore Institute of Technology, Singapore
3
0
6
Hierarchical local attention in TextNCA reveals that the arrangement of attention windows can dramatically influence language modeling performance, even more than the model's iterative nature.
A comprehensive taxonomy reveals critical failure modes in LLM reasoning, exposing vulnerabilities that could hinder their deployment in real-world applications.
Scaling up LLMs doesn't uniformly improve context handling; instead, it paradoxically amplifies the tendency to copy irrelevant tokens while simultaneously improving resistance to misinformation.
