LEV: Thinking Fast and Deep with Looped Decision Models
arXiv Preprint
Dingjie Song*, Zhiling Yan*, Yuxuan Zhang, Lichao Sun
* denotes equal contribution.
TL;DR LEV is a decision model that thinks deeper by rerunning its layers in a loop instead of generating tokens, so a single checkpoint can answer at any depth; a 2.6B LEV goes from 63.2% to 76.2% on JevBench, and on-demand depth saves 15% to 24% of the computation.