Search papers, labs, and topics across Lattice.
6
0
8
MemoryCard transforms long videos into coherent, topic-focused segments, boosting long-video QA accuracy by over 21% while maintaining visual-token efficiency.
An open-source 8B model can now rival proprietary LLMs on a range of complex chemistry tasks by plugging in modules that give it explicit molecular "cognition".
Forget flat retrieval: NaviRAG's hierarchical knowledge navigation boosts RAG performance by dynamically retrieving information across different levels of granularity.
Forget global context – ReAlign leverages a stronger VLM to generate *local*, reasoning-guided descriptions that boost visual document retrieval by up to 2%.
LLMs can be made significantly more reliable in scientific domains by automatically converting flexible knowledge into hard constraints during decoding, boosting accuracy by 12% on average.
By decoupling patch details from semantics, Cheers achieves state-of-the-art multimodal performance at 20% of the training cost of comparable models.
