Search papers, labs, and topics across Lattice.
3
0
7
7
VLMs suffer from "digital agnosia," exhibiting a surprisingly sharp failure to transcribe even small color grids into matrices, revealing a critical gap between visual feature encoding and language generation.
Robots can now assemble complex furniture in zero-shot settings by actively retrieving and reasoning over visual instructions, outperforming methods relying solely on internal knowledge or limited examples.
Even the best LLMs struggle to maintain correct intermediate states when solving university-level STEM problems, often taking more steps than necessary and accumulating errors along the way.
