HorizonBench generates 6-month conversation histories from structured mental state graphs to test AI models on tracking evolving user preferences, finding that frontier models mostly fail at belief updates and perform near or below chance.
arXiv preprint arXiv:2503.07018 , year=
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CL 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
LMEB benchmark shows that embedding models' performance on traditional retrieval does not transfer to long-horizon memory tasks, larger models do not always perform better, and LMEB measures capabilities orthogonal to MTEB.
citing papers explorer
-
HorizonBench: Long-Horizon Personalization with Evolving Preferences
HorizonBench generates 6-month conversation histories from structured mental state graphs to test AI models on tracking evolving user preferences, finding that frontier models mostly fail at belief updates and perform near or below chance.
-
LMEB: Long-horizon Memory Embedding Benchmark
LMEB benchmark shows that embedding models' performance on traditional retrieval does not transfer to long-horizon memory tasks, larger models do not always perform better, and LMEB measures capabilities orthogonal to MTEB.