🏢 Shanghai Artificial Intelligence Laboratory

Align LLMs efficiently via test-time search using smaller models!

AdaptiveDiffusion accelerates diffusion model inference by adaptively skipping noise prediction steps, achieving 2-5x speedup without quality loss.

ProSST, a novel protein language model, integrates protein sequences and structures using quantized structure representation and disentangled attention, achieving state-of-the-art performance in zero-…

MxDNA: Model learns optimal DNA tokenization via gradient descent, outperforming existing methods.

MindMerger efficiently boosts LLM reasoning in non-English languages by merging LLMs with external multilingual language understanding capabilities, achieving significant accuracy improvements, especi…

InternLM-XComposer2-4KHD pioneers high-resolution image understanding in LVLMs, scaling processing from 336 pixels to 4K HD and beyond, achieving state-of-the-art results on multiple benchmarks.

GSDF: A novel dual-branch neural scene representation elegantly resolves the rendering-reconstruction trade-off by synergistically combining 3D Gaussian Splatting and Signed Distance Fields via mutual…

Director3D generates realistic 3D scenes and camera trajectories from text descriptions using a three-stage pipeline: Cinematographer, Decorator, and Detailer.