Search papers, labs, and topics across Lattice.
Affiliation:
3
0
4
Structural attacks on GNNs can be systematically neutralized without retraining simply by pruning the specific edges that disproportionately inflate the graph's kernel complexity.
EAPO revolutionizes LLM reasoning by dynamically integrating prior experiences, leading to consistent performance gains over traditional RLVR methods.
Rigid reward clipping throws away valuable information just beyond the boundary, but a simple stochastic rescue of these signals can substantially boost RLVR performance.
