[
    {
        "id": "osp-23328",
        "type": "article-journal",
        "title": "MemoryAthena: Adaptive Routing over Latent and Generated Memories",
        "author": [
            {
                "family": "Li",
                "given": "Mingyuan"
            },
            {
                "family": "Yu",
                "given": "Guangsheng"
            },
            {
                "family": "Zhang",
                "given": "Juyuan"
            },
            {
                "family": "Wang",
                "given": "Xu"
            },
            {
                "family": "Man",
                "given": "Zhibo"
            },
            {
                "family": "Zhang",
                "given": "Haonan"
            },
            {
                "family": "Ji",
                "given": "Shaoxiong"
            }
        ],
        "URL": "https://omanscience.com/en/articles/memoryathena-adaptive-routing-over-latent-and-generated-memories",
        "language": "en",
        "issued": {
            "date-parts": [
                [
                    2026
                ]
            ]
        },
        "abstract": "Learned-memory methods store information in an explicit table and consume it through a separate reader, allowing addressing, storage, and reading to be modified independently. We study whether useful memory can also be generated rather than only retrieved. MemoryAthena uses three pathways: direct Engram retrieval (E), generation from retrieved Engram cues (GE), and generation from causal backbone states without consulting the memory table (GH). Generated memory is conditionally useful: it can complement E in one context but interfere with it in another. MemoryAthena therefore treats E as an anchor and learns when a generated representation should intervene. With the backbone, memory, generators, and readers frozen, a lightweight causal routing head is trained from counterfactual future-token likelihood advantages of GE and GH relative to E. At inference time, an admitted candidate modifies the E residual through bounded interpolation, while rejection recovers the direct pathway exactly. On question answering, MemoryAthena raises the five-task average from 37.65 to 39.28 over the direct pathway of the same checkpoint, while the six-task general-NLP average increases from 76.73 to 79.13. The complete memory-side system contains approximately 201M parameters, excluding the frozen backbone. Further analyses show complementary strengths among E, GE, and GH across tasks and inputs. These results support generated memory as a selective correction to direct retrieval and highlight routing when, which, and how strongly to intervene as the central challenge."
    }
]