# ExplicitLM: Decoupling Knowledge from Parameters via Explicit Memory Banks

arXiv: https://arxiv.org/abs/2511.01581 · HTML: https://mlsanity.com/paper/2511.01581/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 26 of its 30 references that are in the map.
- First posted on arXiv 3 Nov 2025.

## Builds on (most cited first, 26 of 26)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://mlsanity.com/paper/2312.00752/) (2023)
- [Neural Discrete Representation Learning](https://mlsanity.com/paper/1711.00937/) (2017)
- [Locating and Editing Factual Associations in GPT](https://mlsanity.com/paper/2202.05262/) (2022)
- [Mixtral of Experts](https://mlsanity.com/paper/2401.04088/) (2024)
- [Transformer Feed-Forward Layers Are Key-Value Memories](https://mlsanity.com/paper/2012.14913/) (2020)
- [RWKV: Reinventing RNNs for the Transformer Era](https://mlsanity.com/paper/2305.13048/) (2023)
- [Knowledge Neurons in Pretrained Transformers](https://mlsanity.com/paper/2104.08696/) (2021)
- [Memory-Based Model Editing at Scale](https://mlsanity.com/paper/2206.06520/) (2022)
- [Model Context Protocol (MCP): Landscape, Security Threats, and Future Research Directions](https://mlsanity.com/paper/2503.23278/) (2025)
- [Large Language Models with Controllable Working Memory](https://mlsanity.com/paper/2211.05110/) (2022)
- [Calibrating Factual Knowledge in Pretrained Language Models](https://mlsanity.com/paper/2210.03329/) (2022)
- [Evaluating Retrieval Quality in Retrieval-Augmented Generation](https://mlsanity.com/paper/2404.13781/) (2024)
- [Mixture of A Million Experts](https://mlsanity.com/paper/2407.04153/) (2024)
- [Can LMs Learn New Entities from Descriptions? Challenges in Propagating Injected Knowledge](https://mlsanity.com/paper/2305.01651/) (2023)
- [Dated Data: Tracing Knowledge Cutoffs in Large Language Models](https://mlsanity.com/paper/2403.12958/) (2024)
- [DyKnow: Dynamically Verifying Time-Sensitive Factual Knowledge in LLMs](https://mlsanity.com/paper/2404.08700/) (2024)
- [We Urgently Need Privilege Management in MCP: A Measurement of API Usage in MCP Ecosystems](https://mlsanity.com/paper/2507.06250/) (2025)
- [A Review of the Challenges with Massive Web-mined Corpora Used in Large Language Models Pre-Training](https://mlsanity.com/paper/2407.07630/) (2024)
- [Dual Cognitive Architecture: Incorporating Biases and Multi-Memory Systems for Lifelong Learning](https://mlsanity.com/paper/2310.11341/) (2023)

## Built on by (most cited first, 1 of 1)

- [Memory for Large Language Models](https://mlsanity.com/paper/2607.25380/) (2026)

## Abstract (arXiv)

Large language models suffer from knowledge staleness and lack of interpretability due to implicit knowledge storage across entangled network parameters, preventing targeted updates and reasoning transparency. We propose ExplicitLM, a novel architecture featuring a million-scale external memory bank storing human-readable knowledge as token sequences, enabling direct inspection and modification. We design a differentiable two-stage retrieval mechanism with efficient coarse-grained filtering via product key decomposition (reducing complexity from O(N · |I|) to O(√N · |I|)) and fine-grained Gumbel-Softmax matching for end-to-end training. Inspired by dual-system cognitive theory, we partition knowledge into frozen explicit facts (20%) and learnable implicit patterns (80%), maintained through Exponential Moving Average updates for stability. ExplicitLM achieves up to 43.67% improvement on knowledge-intensive tasks versus standard Transformers, with 3.62× gains in low-data regimes (10k samples). Analysis shows strong correlations between memory retrieval and performance, with correct predictions achieving 49% higher hit rates. Unlike RAG systems with frozen retrieval, our jointly optimized architecture demonstrates that interpretable, updatable models can maintain competitive performance while providing unprecedented knowledge transparency.