A training-inference consistent segmented execution framework for long-context LLMs matches full-context performance with substantially lower peak memory at very long lengths.
Lm-infinite: Zero-shot extreme length generalization for large language models
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
representative citing papers
SharedLLM stacks two copies of a short-context LLM so the lower one compresses context into query-aware multi-grained tokens that are injected only at the lowest layers of the upper one, enabling generalization from 8K training to 128K+ inputs.
Titans combine attention for current context with a learnable neural memory for long-term history, achieving better performance and scaling to over 2M-token contexts on language, reasoning, genomics, and time-series tasks.
citing papers explorer
-
Training-Inference Consistent Segmented Execution for Long-Context LLMs
A training-inference consistent segmented execution framework for long-context LLMs matches full-context performance with substantially lower peak memory at very long lengths.
-
Stacked from One: Multi-Scale Self-Injection for Context Window Extension
SharedLLM stacks two copies of a short-context LLM so the lower one compresses context into query-aware multi-grained tokens that are injected only at the lowest layers of the upper one, enabling generalization from 8K training to 128K+ inputs.
-
Titans: Learning to Memorize at Test Time
Titans combine attention for current context with a learnable neural memory for long-term history, achieving better performance and scaling to over 2M-token contexts on language, reasoning, genomics, and time-series tasks.