Search papers, labs, and topics across Lattice.
3
1
4
7
Language models are increasingly doing their real work in the "invisible" latent space, not the tokens we see.
By explicitly modeling and calibrating a model's intrinsic uncertainty, EGPO unlocks significant gains in reasoning performance for RL-trained language models.
LLMs can now dynamically create and refine their own scientific tools at test time, outperforming agents stuck with static toolsets.
