Search papers, labs, and topics across Lattice.
4
0
5
18
Current reliance metrics miss the crucial shift in cognitive effort distribution between users and AI, but this new "offloading score" finally captures how much work is *actually* being delegated to the tool.
Turns out, coding agents in the wild are only writing useful code 44% of the time, and are introducing more security vulnerabilities than human developers.
Turns out, ensembling LLMs for diversity isn't about picking the biggest model, but routing each prompt to the *right* model.
LLMs can now conduct better qualitative interviews than ever before, adaptively balancing structured topic coverage with the serendipitous discovery of novel insights.
