KRAFTON
Researchers Introduce Meta-Harness Tool to Boost LLM Accuracy
Researchers from Stanford, MIT, and KRAFTON have built Meta-Harness, a tool boosting LLM accuracy by 7.7 points with 4× fewer context tokens.
Researchers from Stanford, MIT, and KRAFTON have built Meta-Harness, a tool boosting LLM accuracy by 7.7 points with 4× fewer context tokens.