Research2h ago
RENDER Benchmark Tests LLM Memory Formatting
Matched-budget resolved packets outperformed recency-truncated raw dialogue by 42.4 to 72.6 points in the new RENDER benchmark. The study tests how different input formats, like LangChain summaries or raw excerpts, affect memory retrieval. Results show a performance spread of up to 48.8 points based solely on how evidence is rendered.