Pith. sign in

REVIEW 6 cited by

M+: Extending MemoryLLM with Scalable Long-Term Memory

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2502.00592 v2 pith:W6RE74OE submitted 2025-02-01 cs.CL

classification cs.CL
keywords memorymemoryllminformationknowledgelong-termretentiontokensextending
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Equipping large language models (LLMs) with latent-space memory has attracted increasing attention as they can extend the context window of existing language models. However, retaining information from the distant past remains a challenge. For example, MemoryLLM (Wang et al., 2024a), as a representative work with latent-space memory, compresses past information into hidden states across all layers, forming a memory pool of 1B parameters. While effective for sequence lengths up to 16k tokens, it struggles to retain knowledge beyond 20k tokens. In this work, we address this limitation by introducing M+, a memory-augmented model based on MemoryLLM that significantly enhances long-term information retention. M+ integrates a long-term memory mechanism with a co-trained retriever, dynamically retrieving relevant information during text generation. We evaluate M+ on diverse benchmarks, including long-context understanding and knowledge retention tasks. Experimental results show that M+ significantly outperforms MemoryLLM and recent strong baselines, extending knowledge retention from under 20k to over 160k tokens with similar GPU memory overhead. We open-source our code at https://github.com/wangyu-ustc/MemoryLLM

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Light-Omni: Reflex over Reasoning in Agentic Video Understanding with Long-Term Memory

    cs.CV 2026-07 conditional novelty 6.5 of 10

    Dual global+latent states with hierarchical episodic merging enable reflexive, low-latency long-video agents that beat iterative reasoning baselines on accuracy and efficiency.

  2. A-TMA: Decoupling State-Aware Memory Failures in Long-Term Agent Memory

    cs.AI 2026-07 unverdicted novelty 6.0 of 10

    Explicit bank, retrieval, and QA state roles reduce ghost-memory failures on conflict-heavy LTP and improve some temporal scores on LoCoMo, with host-dependent gains.

  3. Towards General Continuous Memory for Vision-Language Models

    cs.LG 2025-05 conditional novelty 5.0 of 10

    A vision-language model can act as its own continuous memory encoder, compressing external multimodal knowledge into eight embeddings that improve reasoning when prepended to the frozen model.

  4. Narrative Memory in Machines: Multi-Agent Arc Extraction in Serialized TV

    cs.MM 2025-08 conditional novelty 4.0 of 10

    A multi-agent LLM system with a vector database extracts narrative arcs from TV episode summaries, scoring 89% precision on self-contained arcs while missing overlapping relationship arcs.

  5. Token-Operations-Oriented Inference Optimization Techniques for Large Models

    cs.SE 2026-06 unverdicted novelty 3.0 of 10

    The paper introduces a four-layer technical architecture for token-operations-oriented inference optimization in large models and reviews key technologies and industry status at each layer.

  6. Memory-Augmented Transformers: A Systematic Review from Neuroscience Principles to Enhanced Model Architectures

    cs.LG 2025-08 unverdicted novelty 3.0 of 10

    Memory-augmented Transformer research is organized into a three-axis taxonomy bridging neuroscience memory concepts to network designs, but no new result is produced.

Pith tools