Author

O. Koyejo

2 papers indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#artificial intelligence Sep 2026

Mind the Spike: Mechanisms and Brittleness of Visual Massive Activations in Large Vision-Language Models

Large vision-language models (LVLMs) inherit massive activations from their text-only bases: spikes where a few fixed hidden channels receive values thousands of times above the typical magnitude. The text spike systematically appears in early layers at a fixed initial position, independently of input content. Visual s...

Jonas Ngnawé, Y. Pequignot, Sabyasachi Sahoo et al. · arXiv.org · 0 citations

#artificial intelligence Conference Feb 2025

Logits are All We Need to Adapt Closed Models

A token-level probability reweighting framework that, given access to logits and a small amount of task-specific data, can effectively steer black-box LLMs toward application-specific content generation and advocates for broader access to token logits in closed-source models.

G. Hiranandani, Hao-Lun Wu, Subhojyoti Mukherjee et al. · International Conference on... · 7 citations · ⚡1