Search papers, labs, and topics across Lattice.
Technical University of Munich
3
0
5
Log$_\text{b}$Quant achieves superior quantization performance, enabling efficient deployment of language models on consumer hardware without sacrificing accuracy.
Finally, a way to measure the "zoom level" of text without needing a reference, opening up new possibilities for understanding model behavior and dataset difficulty in QA.
Unleash the hidden potential of LLM reasoning with Framework of Thoughts, a new framework that dynamically optimizes reasoning structures like ToT and GoT for speed, cost, and accuracy.
