Search papers, labs, and topics across Lattice.
Affiliation:
5
0
7
4
A subtle "question-relay" mechanism, not direct generation pathways, is the primary driver of AVLLM hallucinations, and steering these question states can effectively mitigate them.
A novel Dual-Latent Space Reinforcement Learning (DLSRL) framework, which complements initial-noise steering with representation-level control inside the frozen generator and effectively accelerates online robot policy adaptation and achieves competitive performance.
Stateagent transforms video generation by effectively tracking and inferring world states, leading to a 54% improvement in coherence across segments.
AudioLens-R1 redefines audio clustering by allowing models to adaptively organize speech based on user-specified perspectives, achieving unprecedented accuracy improvements.
Realistic predictions in robotic manipulation now come with precise arm control, ensuring the right actions yield the expected outcomes.
