Search papers, labs, and topics across Lattice.
5
0
10
2
Achieve 8x token reduction in million-token document understanding without sacrificing accuracy by having the LLM actively search for relevant information like a foraging animal.
LLMs can edit code 30% faster and cheaper without sacrificing accuracy, simply by learning to choose between generating full code and structure-aware diffs.
Forget querying a single database – Blue's Data Intelligence Layer treats LLMs, the web, and even the user as queryable "databases" to answer complex, real-world questions.
Naturalness-based data selection, a common technique for curating LLM reasoning datasets, systematically favors longer, lower-quality reasoning chains due to a previously unnoticed "step length confounding" effect.
Training domain-specific coding LLMs with realistic environments and large-scale RL can yield substantial gains in practical software engineering tasks.
