CLAM compresses a session stream into a low-rank adapter: no gradients, no generated tokens, and memory built 3–151× faster than extraction systems
CLAM (Carried Latent Adapter Memory) writes long-term memory into model weights: a recurrent hypernetwork compiles a whole stream of session activations into a low-rank adapter in a single forward pass, paired with a verbatim retrieval buffer. On LoCoMo and LongMemEval-S across three open 2–7B base models, it leads all agentic memory baselines by 1–45 points and all gradient-based parametric methods by 10–45 points, building its memory with zero generative calls, zero gradient steps, and 3–151× less wall-clock time than extraction systems. The two channels are complementary: the adapter still adds 12–15 points when the buffer misses the gold evidence.