Search papers, labs, and topics across Lattice.
Shanghai Jiao Tong University
3
0
6
Transforming continuous latent states into discrete tokens enables up to 20脳 compression while enhancing interpretability in reasoning tasks.
Agents can now escape the Self-Confirmation Trap, leading to more reliable experience learning and improved self-evolution.
Forget task-specific overfitting: training coding agents on atomic skills unlocks surprisingly broad generalization to complex software engineering tasks.
