Search papers, labs, and topics across Lattice.
2
0
6
6
State-of-the-art LLM agents face a staggering performance decline in multilingual workflows, revealing critical gaps in current evaluation methods.
Latent visual reasoning in multimodal LLMs is largely ineffective, as the "imagination" happening in latent space doesn't actually attend to the input or influence the output, making explicit text-based imagination a surprisingly better alternative.
