Proteus: Incremental Memory Activation for Long-Context Sequence Modeling
The quadratic cost of attention-based sequence models for long contexts has motivated a growing line of research on memory-based models that can compress context into a compact state. However, most existing memory models expose a static memory throughout the entire sequence. Because early tokens face no compression pressure, they occupy too many degrees of freedom and "pollute" the memory state, leaving little capacity for later context and increasing interference between what is stored and what arrives next. We study a new paradigm of incremental memory activation, where the effective capacit
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 67%Breakthrough in long-context efficiency announced →
- PossiblePossibly related (embedding) · 52%New Server Hopes to Break Through AI’s “Memory Wall” →
- LinkedLinked via arxiv author · 85%Reza Bayat →
“Proteus: Incremental Memory Activation for Long-Context Sequence Modeling”
- LinkedLinked via arxiv author · 85%Ali Behrouz →
“Proteus: Incremental Memory Activation for Long-Context Sequence Modeling”
- LinkedLinked via arxiv author · 85%Vahab Mirrokni →
“Proteus: Incremental Memory Activation for Long-Context Sequence Modeling”
- LinkedLinked via arxiv author · 85%Aaron Courville →
“Proteus: Incremental Memory Activation for Long-Context Sequence Modeling”