Search papers, labs, and topics across Lattice.
3
0
7
2
Multimodal models can now achieve state-of-the-art performance in real-world tasks like document understanding and audio-video comprehension with significantly reduced inference latency thanks to novel token-reduction techniques.
LLMs can achieve 2.5x higher throughput and 10.7x KV memory reduction in long-context reasoning by compressing the KV cache using trigonometric functions derived from pre-RoPE query/key vector distributions.
By cleverly repurposing an unused sign bit, IF4 achieves superior quantization performance compared to NVFP4 without increasing bit-width.
