Search papers, labs, and topics across Lattice.
Affiliation:
4
0
7
3
Visually plausible charts generated by LLMs often mask significant data-level hallucinations, revealing a critical gap in current AI capabilities.
AVTP achieves a 2x inference speedup in LVLMs while maintaining up to 96.1% accuracy, revolutionizing multi-image processing efficiency.
LLMs get a reasoning boost by treating information extraction not as a one-off task, but as a dynamic cache that persists and filters information across multiple steps.
GPTQ-style quantization can be significantly improved by directly aligning quantized layer outputs with the original full-precision model's output, rather than the compensated weights, and accounting for "compensation-aware error."
