The authors formulate context adaptation as an optimization problem in which an agent updates textual memory across examples.
Graph memory keeps continual LLM adaptation efficient as experience accumulates
GraphMemory retrieves small, relevant clusters of reusable strategies instead of placing an agent’s entire accumulated memory into every prompt.
Big Tech
Yehya Farhat · Michael Desmond · Anastasios Kyrillidis
Rice University · IBM Research
Research Digest··2 min read
Farhat, Desmond and Kyrillidis present GraphMemory, an external memory that lets frozen language models learn from successive interactions without updating model weights.
Why this paper
From IBM Research and Rice University
In one line
GraphMemory stores reusable strategies as a graph and retrieves only relevant subgraphs, keeping memory-context length constant and reducing token usage.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ✓Limitations stated by the authors (3 noted)
- ✓Reports numbers on named benchmarks (2 benchmarks)
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§