mlsanityData to 14 Oct 2026

Memorizing is Not Enough: Deep Knowledge Injection Through Reasoning

arXiv 2504.00472 ↗First posted 1 Apr 2025cs.CL cs.AIcited by 4

On the map: LLM reasoning → Reasoning benchmarks and evaluation

Abstract (arXiv)

Although large language models (LLMs) excel in knowledge recall and reasoning, their static nature leads to outdated information as the real world evolves or when adapting to domain-specific knowledge, highlighting the need for effective knowledge injection. However, current research on knowledge injection remains superficial, mainly focusing on knowledge memorization and retrieval. This paper proposes a four-tier knowledge injection framework that systematically defines the levels of knowledge injection: memorization, retrieval, reasoning, and association. Based on this framework, we introduce DeepKnowledge, a synthetic experimental testbed designed for fine-grained evaluation of the depth of knowledge injection across three knowledge types (novel, incremental, and updated). We then explore various knowledge injection scenarios and evaluate the depth of knowledge injection for each scenario on the benchmark. Experimental results reveal key factors to reach each level of knowledge injection for LLMs and establish a mapping between the levels of knowledge injection and the corresponding suitable injection methods, aiming to provide a comprehensive approach for efficient knowledge injection across various levels.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning. Builds on 11 of its 23 references that are in the map. First posted on arXiv 1 Apr 2025.

Builds on

11 of 23 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024Grokked Transformers are Implicit Reasoners: A Mechanistic Journey to the Edge of Generalization23 May

Other references · 10

  1. 2024Does Fine-Tuning LLMs on New Knowledge Encourage Hallucinations?cited by 232
  2. 2023Fine-Tuning or Retrieval? Comparing Knowledge Injection in LLMscited by 178
  3. 2024A Comprehensive Study of Knowledge Editing for Large Language Modelscited by 163
  4. 2022A Review on Language Models as Knowledge Basescited by 157
  5. 2023Physics of Language Models: Part 3.2, Knowledge Manipulationcited by 143
  6. 2024How Do Large Language Models Acquire Factual Knowledge During Pretraining?cited by 90
  7. 2024Instruction-tuned Language Models are Better Knowledge Learnerscited by 55
  8. 2023How Do Large Language Models Capture the Ever-changing World Knowledge? A Review of Recent Advancescited by 33
  9. 2024Beyond Memorization: The Challenge of Random Memory Access in Language Modelscited by 8
  10. 2021A Survey of Knowledge Enhanced Pre-trained Modelscited by 3

Built on by

Cited by year

Citing papers per year: 2025: 4.

4 papers.

  1. 2025How do Transformers Learn Implicit Reasoning?cited by 16
  2. 2025Bidirectional LMs are Better Knowledge Memorizers? A Benchmark for Real-world Knowledge Injectioncited by 7
  3. 2025Rote Learning Considered Useful: Generalizing over Memorized Data in LLMscited by 5
  4. 2025PropMEND: Hypernetworks for Knowledge Propagation in LLMscited by 3

Similar, not cited

Close in wording, no citation either way

  1. 2025Knowledge-Instruct: Effective Continual Pre-training from Limited Data using Instructions
  2. 2025Memorization vs. Reasoning: Updating LLMs with New Knowledge
  3. 2024Synthetic Knowledge Ingestion: Towards Knowledge Refinement and Injection for Enhancing Large Language Models
  4. 2025Comparing Knowledge Injection Methods for LLMs in a Low-Resource Regime
  5. 2024Disentangling Memory and Reasoning Ability in Large Language Models
  6. 2025Decoupling Reasoning and Knowledge Injection for In-Context Knowledge Editing
  7. 2024DeepEdit: Knowledge Editing as Decoding with Constraints
  8. 2022KMIR: A Benchmark for Evaluating Knowledge Memorization, Identification and Reasoning Abilities of Language Models