Authors’ Reply: Citation Accuracy Challenges Posed by Large Language Models
Bibliographic record
Abstract
Large language models (LLMs) have demonstrated significant potential in academic research but face challenges in generating accurate citations. The issue of hallucinated references—well-formatted but fictitious citations—arises due to LLMs' limited access to subscription-based databases and their reliance on probabilistic text generation. This letter discusses two key approaches to mitigating these issues. First, retrieval-augmented generation (RAG) combined with Hallucination Aware Tuning (HAT) improves citation integrity by integrating external databases and employing hallucination detection models. However, even RAG-HAT systems may still misinterpret source content. Second, we propose the development of “Reference-Accurate” Academic LLMs by major global publishers, which would be trained exclusively on rigorously verified academic literature, ensuring that all citations generated are authentic and traceable. We recommend a dual approach integrating RAG-HAT with publisher-backed academic LLMs, along with human oversight, to enhance AI-assisted scholarly communication. Future research should evaluate the accuracy and reliability of these methods to promote responsible AI use in academia.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.009 | 0.117 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.003 | 0.003 |
| Scholarly communication | 0.004 | 0.005 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.019 | 0.027 |
| Insufficient payload (model declined to judge) | 0.016 | 0.012 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".