Search papers, labs, and topics across Lattice.
4
0
9
1
AVA-Encoder achieves a 73.1% relative improvement in video representation learning, enabling agents to produce cinematic-grade videos with far fewer resources.
Forget training separate interventions – UniSteer lets you steer LLMs across diverse behaviors, from truthfulness to persona, using a single text-guided flow in activation space.
LLMs can now forecast time series data with greater accuracy and interpretability by fusing their reasoning capabilities with traditional time series models in a novel agentic framework.
Unlock more diverse and effective LLM outputs by explicitly rewarding semantic novelty during decoding with Exploratory Sampling.
