Digital Repository
Belarusian state university of foreign languages
ISSN (online):
Please use this identifier to cite or link to this item:
http://e-lib.bsufl.by/handle/edoc/17819| Title: | Restoring scientific memory: regenerative reception of concepts via large language models and neuro-logical integration |
| Authors: | Kimiaki Tokumaru |
| Keywords: | Языкознание (лингвистика) Прикладное языкознание Информатика и информационные технологии |
| Issue Date: | 2025 |
| Publisher: | Белорусский государственный университет иностранных языков |
| Citation: | Kimiaki Tokumaru. Restoring scientific memory: regenerative reception of concepts via large language models and neuro-logical integration / Kimiaki Tokumaru // Язык в эпоху цифровых трансформаций и развития искусственного интеллекта : сб. науч. ст. по итогам II междунар. науч. конф., Минск, 23–24 окт. 2025 г. / редкол. : Н. Е. Лаптева (отв. ред.) [и др.]. – Минск : БГУИЯ, 2025. – С. 261–267. |
| Description: | Scientific knowledge is not merely accumulated – it is biologically and logically regenerated. This paper proposes a model of regenerative reception for restoring fragmented scientific concepts, integrating three key components: (1) the biological substrate of conceptual memory, hypothesized to reside in cerebrospinal fluid (CSF)-resident B-cells; (2) the logical distinction between semantic and episodic memory in scientific versus everyday concept formation; and (3) the instrumental role of Large Language Models (LLMs) in detecting contradictions at the physical layer of language. We argue that LLMs, while incapable of semantic integration, can reliably identify terminological and structural inconsistencies in scientific texts. These contradictions, once flagged, must be addressed through a human-led regenerative process grounded in primary sources and the logical architectures established by original discoverers. Interpretation is explicitly prohibited until full epistemic reconstruction is achieved. The result is not a static archive but a regenerative ecology of scientific memory, where meaning is not transmitted but reactivated within the biological consciousness of the learner. Key words: regenerative reception; scientific memory; B-lymphocytes in CSF; contradiction detection; physical vs logical layer; large language models (LLMs), semantic reconstruction. |
| URI: | http://e-lib.bsufl.by/handle/edoc/17819 |
| ISBN: | 978-985-28-0275-8 |
| Appears in Collections: | Язык в эпоху цифровых трансформаций и развития искусственного интеллекта (статьи) |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 261-267.pdf | 7,99 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.