New benchmark tests how well small language models retain long-context memory
Researchers introduced SRRM, a retrieval-reconstruction benchmark on arXiv that evaluates how small language models preserve information across progressively longer contexts when deployed on consumer-grade hardware.
Read the full story at arXiv cs.CL ↗
Timeline · 1 report
- 2026-11-05 05:00 · arXiv cs.CL
SRRM: A Retrieval-Reconstruction Benchmark for Evaluating Long-Context Memory in Small Language Models