How we took LongMemEval from 80.0 to 94.2 without touching retrieval
LongMemEval-S climbed from 80.0 to 94.2 while retrieval recall at 10 stayed at 0.99. The gains came from evidence budget, reader generation, and instructions.
Tagged
LongMemEval-S climbed from 80.0 to 94.2 while retrieval recall at 10 stayed at 0.99. The gains came from evidence budget, reader generation, and instructions.