Benchmark

StuLife

An academic benchmark designed to evaluate agents' long-range information recall ability across multiple rounds of interaction, focusing on how well agents retain and retrieve information over extended multi-turn dialogues.

Timeline (last 90 days)

Sep 24

On the recall-focused portion of the StuLife benchmark, JAZ's invoke outperforms Letta (MemGPT) by 8% at half the cost.

Unverified50%

Sep 24

StuLife is an academic benchmark focused on evaluating an agent's ability to recall long-range information across multiple rounds of interaction.

Unverified50%

All Facts (2)

Unverified

On the recall-focused portion of the StuLife benchmark, JAZ's invoke outperforms Letta (MemGPT) by 8% at half the cost.

50%Unverified

StuLife is an academic benchmark focused on evaluating an agent's ability to recall long-range information across multiple rounds of interaction.

50%

Source Articles

JAZ Framework: Rethinking LLM Agent Design with a Minimal invoke PrimitiveSep 24