Search papers, labs, and topics across Lattice.
kpd@illinois.edu, CSAIL/LIDS, Massachusetts Institute of Technology & Siebel School of Computing and Data Science, University of Illinois, Urbana-Champaignmragh@mit.edu, MIT Sloan School of Management & Department of Electrical Engineering and Computer Science
MIT CSAIL1
1
2
1
Standard winrate metrics in LLM evaluation can backfire, incentivizing model creators to produce homogenous models that actually *decrease* overall consumer welfare.
