benchmark
Multi-hop Graph Retrieval
benchmarkactiveprovisional
multi-hop-graph-retrieval-aa54b3a3·1 events·first seen 22d agoAliases: Multi-hop Graph Retrieval
Co-occurring entities
More like this (12)
Multi-hop Question Answeringmulti-hop reasoningGraph Neural Network leveraging Higher-order Class Label Connectivity for Heterophilous GraphsKnowledge Graph Random Walkheterogeneous graph memorydual-graph frameworkGraph Structure LearningHierarchical Graph Memoryhybrid dense-sparse retrievalcentroid-guided routingLangGraphGraphWalks
Recent events (1)
Language Models Need Sleep: Periodic Context Consolidation via Fast Weights and SSM Blocks
This paper proposes a sleep-like consolidation mechanism for transformer-based LLMs to address the quadratic scaling of attention with context length. During 'sleep' phases, the model performs N offline recurrent passes over accumulated context, updating fast weights in state-space model (SSM) blocks via a learned local rule, then clears the KV cache. The approach is evaluated on synthetic tasks (cellular automata, multi-hop graph retrieval) and math reasoning, where standard transformers and SSM-attention hybrids fail, with performance scaling with sleep duration N.