Search papers, labs, and topics across Lattice.
4
0
8
13
Autoresearch agents can waste compute resolving the same issues repeatedly, but targeted interventions can dramatically enhance their efficiency and performance.
By rethinking hypothesis generation as a sampling problem, this framework boosts both the quality and diversity of scientific hypotheses, challenging the status quo of optimization-focused methods.
Multi-modal alignment in symbolic regression models like SNIP doesn't actually improve during optimization, suggesting current approaches are too coarse to effectively guide symbolic search.
Tabular foundation models can now tackle multivariate time series forecasting in a zero-shot manner, outperforming existing tabular methods by cleverly reframing the problem as a series of scalar regressions.
