Search papers, labs, and topics across Lattice.
3
0
7
Mid-tier LLMs outperform their stronger counterparts in harness self-evolution, challenging assumptions about model capability and adaptability.
Achieve 50% bitrate savings in ultra-low-bitrate image compression by cleverly turning image decoding into a next-frame prediction problem using video diffusion priors.
Latent reasoning models often take shortcuts to achieve high accuracy, and stronger supervision, while mitigating this, paradoxically restricts the diversity of their latent representations.
