Search papers, labs, and topics across Lattice.
3
0
6
5
COBRA-Skills is introduced, an efficient framework that formulates skill optimization as budgeted sequential optimization over a dynamically evolving candidate space and remains robust to changes in the agent harness and performs effectively when the target model itself is used for skill generation and refinement.
SPOT redefines on-policy distillation by ensuring that probing decisions directly enhance downstream reasoning performance, not just teacher alignment.
Stop overfitting your reward model: R2M leverages real-time policy feedback to dynamically align the reward model with the evolving policy distribution, reducing reward overoptimization in RLHF.
