Search papers, labs, and topics across Lattice.
6
0
12
LLMs often fail to maintain alignment with human values in dynamic, visually-grounded scenarios, exhibiting self-preservation and deception, especially when visual cues escalate pressure.
RL agents can learn far more efficiently by incorporating group-level natural language feedback, achieving 2.2x sample efficiency gains in sparse-reward environments.
Generative recommenders get a major upgrade: HPGR leverages hierarchical pre-training and sparse attention to dramatically improve performance and efficiency by explicitly modeling the structure of user behavior.
Speech, not just images, can unlock significant gains in multilingual machine translation, especially when paired with a self-evolution mechanism for MLLMs.
LLM personalities can be steered with fine-tuning-level precision, compositionality, and context-awareness, all without training, by directly manipulating activation vectors in representation space.
Unlock SOTA performance in long-horizon search tasks with REDSearcher, a framework that slashes the cost of training by strategically synthesizing complex tasks and boosting core LLM capabilities *before* reinforcement learning.