Stanford SNAP group (Jure Leskovec), with co-authors from Toronto, Penn and MIT · Latent Memory Module
Memorilla
Memorilla is a 194M-parameter latent memory module that encodes a document collection into 16 memory tokens for a frozen Qwen3-8B decoder, with a frozen Qwen3-Embedding-4B encoder. On FactKG it scored 84.33% against 80.43% with full context, using 108x fewer document tokens, and it reports the highest average (49.68) across eight benchmarks, though it trails MemoRAG on some tasks and PubMedQA. The benchmark scores are self-reported. Code, training recipes and checkpoints are released under Apache 2.0, and the team also released the PersonalizationV4 dataset under CC BY 4.0.
Open the source ↗Similar Research & Safety releases
- NovaAtom-Lite
Geodesic Intelligence · All-atom structure prediction · Week 47, 2026
- Actor, Judge, or Teacher? Understanding Decision Models in Long-Horizon Agents
Haochen Liu, Qianhao Zhu, Zixuan Wang, Dingming Li, Tongbo Chen, Peng Qiu, Yingqi Qu, Weiming Lu (Zhejiang University) · Decision Model Teachers · Week 47, 2026
- PhysLoop
DAGroup-PKU (Yuxuan Chen et al.) · Physics-aware self-improvement · Week 47, 2026
- Human-in-the-loop Topology Optimisation Copilot
MIT (Dat Ha, Josephine Carstensen, Ferdous Alam, Markus Buehler, Faez Ahmed) · Design Copilot · Week 47, 2026
- Read Less, Solve More: Token-Efficient Sparse Reading of External Content for AI Agents
Zedong Liu, Jiaan Wu, Xinyang Ma, Le Xu, Kai Wang, Yuanchao Hu, Dingwen Tao, Guangming Tan (Institute of Computing Technology) · Token-efficient agent reading · Week 47, 2026
- On the Design of Hyper-Connections: Simplification, Stability, and Scalability
Xiangdong Zhang et al. (dots studio) · Simpler Hyper-Connections · Week 47, 2026
Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.