Search papers, labs, and topics across Lattice.
7
0
8
9
Achieving intended outcomes in video generation while maintaining semantic relevance is more challenging than previously thought, with current models falling short.
TreeCredit redefines credit assignment in multi-agent reasoning, leading to better accuracy and lower inference costs through innovative state-matched comparisons.
Existing AI-generated image detectors falter dramatically, with accuracy plummeting from 91-96% to as low as 54-66% when faced with realistic manipulations.
TriMatch achieves superior correspondence discrimination by integrating geometric and semantic features, effectively tackling the issue of pseudo-consistent outliers in complex scenes.
LLMs can now simulate e-commerce dispute resolution more effectively than existing methods, thanks to a multi-agent framework that mimics the deliberation and voting patterns of crowdsourced jurors.
MLLMs can overcome self-referential bias and improve visual grounding by actively exploring and correcting their cognitive deficiencies, guided by token-level epistemic uncertainty.
Stop wasting compute on unreliable rollouts and easy frames: Stream-R1 adaptively focuses video diffusion distillation where it matters most, boosting quality without architectural changes or added inference cost.
