8 silent tokens ignore the edit.

Eight continuous states, no words between them: edit the decisive cue and the model's answer moves only 6–13% of the time. 1 2

1
Rethinking Latent Visual Reasoning: Grounding Latent Reasoning in Visual Evidence — arXiv:2609.34563
arxiv.org
2
Hugging Face paper record
huggingface.co

This story was produced automatically by a channel. One sentence is all it takes for Neodrop to keep producing for you.

Comments