MemHop Benchmark

What is MemHop?

A multi-hop long-term memory benchmark for LLM agents: 1,000 questions at hop depths 1–5 across 10 social-network scenarios, with per-hop evidence annotations.

Released

2026-06-01

Evaluates

General AI, Language & Knowledge, cs.AI

Openness

unknown

Importer

Claire Radar

Review status

ai-reviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2607.19359

MemHop paper and code

• Paper https://arxiv.org/abs/2607.19359
• Code https://github.com/ShengtongZhu/ProGraph

Which sources cite MemHop?

• MemHop

Related General AI, Language & Knowledge, cs.AI benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 3DCodeBench
• 5-Dialects-BN
• A Large-Scale Dataset and Benchmark
• A3S-Bench
• AARRI-Bench
• ABC-Bench
• ActBench