There is a newer version of the record available.

Published September 26, 2026 | Version v0.4.0

Benign Memory Contamination in LLM Agents: A Taxonomy, Benchmark, and Contract-Level Mitigations

Authors/Creators

Description

CONTAM-Bench measures how persistent-memory features degrade LLM assistant responses through benign contamination

Notes

If you use this work, please cite it using the metadata below.

Files

arananet/contam-bench-v0.4.0.zip

Files (1.0 MB)

Name Size
md5:dcb824f1f6eceecaf675e085b74684b3
1.0 MB Preview Download

Additional details

Related works