Agent / long-term memory
AriGraph: Learning Knowledge Graph World Models with Episodic Memory for LLM Agents
Cited as a baseline — critiqued by newer work, not yet beaten on a benchmark here
1 papers critique it · 0 beat it on benchmarks
Verbatim critique sentences, each from a paper that cites AriGraph as a baseline.
tailored for text-based games with discrete state transitions, which are rarely present in the ambiguous boundaries of natural conversation.
Recent methods in the same sub-problem, not yet superseded in the knowledge base — arXiv benchmark leaders, not vetted production recommendations.