Search papers, labs, and topics across Lattice.
Affiliation:
4
0
7
Outperforming prior open models, Instella-MoE sets a new standard for efficiency and performance in language modeling with its innovative architecture and training methods.
Captions selected with VEGAS align significantly better with human attention, boosting retrieval performance and challenging the status quo of video captioning metrics.
Forget training from scratch: HyLo lets you breathe new (long-context) life into your existing Transformer LLMs, achieving 32x context extension and 90% KV-cache reduction.
Diffusion models can be made more efficient by dynamically allocating compute to image regions based on their information content, achieving better image quality with fewer FLOPs.
