Track new results across pure and applied mathematics.
Search papers, labs, and topics across Lattice.
3
0
4
0
LLMs can learn from instructions and tasks entirely outside their pretraining distribution by decomposing them into known subtasks.
Asynchronous OPD can boost training throughput significantly while managing the challenges of stale data, transforming the efficiency of large language model fine-tuning.
Just two factors can explain over 90% of a model's performance across 133 benchmarks, drastically simplifying evaluation processes.