AI citation finder
Paste a written paragraph and every claim that needs a source gets real references. Only citations are inserted; your text stays as it is.
Paste text or upload .docx / .txt / .md. Credits are charged by actual usage. See credit details
How to find references for a paper
- 1
Paste a paragraph
Paste an introduction, related work or any paragraph that needs sources, or upload a file.
- 2
Find citation points
Sentences describing prior work, citing others' data or restating others' views are marked; your own contributions, common knowledge and transitions are not.
- 3
Search and verify
Each spot gets English and Chinese queries searched in rounds; full abstracts are read and only papers that support the sentence are kept.
- 4
Insert citations
Author–year citations linked to the original go before the sentence's end punctuation, and an existing reference list is updated.
Accurate citations, untouched text
Every citation comes from a real search and was checked against its abstract.
Only papers that support the claim
Being on topic is not the same as supporting the sentence. Each spot gets candidates, their full abstracts are read, and only the 1–3 that actually support the claim are kept.
Only sentences that need a source
Sentences on prior work, others' data or others' views are marked. Your own methods and conclusions, common knowledge and transitions are not, and sentences already cited are skipped.
No forced citations
When no supporting paper turns up, nothing is forced in. The report lists those spots so you can rephrase or add keywords and try again.
Citations in, nothing else changed
Citations go just before the end punctuation and no other word of your text changes. If there is already a reference list, new papers are added in the same format.
Watch a paragraph get its references
From finding citation points to inserting citations, each step happens in the workspace. The paragraph is written to show the process; all cited papers are real.
Medical image segmentation is a basic step in computer-aided diagnosis, and deep learning has become the mainstream approach to it (Litjens et al., 2017). Encoder–decoder architectures represented by U-Net perform well and can be trained with only a few annotated images (Ronneberger et al., 2015; Isensee et al., 2021). In recent years, Transformer-based methods have further strengthened the modeling of long-range dependencies (Chen et al., 2021). However, most methods are evaluated on single-center data, and their generalization across centers lacks systematic comparison. This paper proposes an adaptive segmentation framework for cross-center data.
References
- Litjens G, Kooi T, Bejnordi B E, et al. A survey on deep learning in medical image analysis. Medical Image Analysis, 2017.
- Ronneberger O, Fischer P, Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation. MICCAI, 2015.
- Isensee F, Jaeger P F, Kohl S A A, et al. nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nature Methods, 2021.
- Chen J, Lu Y, Yu Q, et al. TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation. arXiv:2102.04306, 2021.
Add references to this paragraph
Checking candidates
- Sentence 2Ronneberger et al., 2015
Abstract: trains end to end from very few images
- Sentence 2Isensee et al., 2021
Abstract: surpasses most approaches on 23 public datasets
- Sentence 2Long et al., 2015
Semantic segmentation of natural images; does not support this sentence
- Sentence 4
None of the results compare across centers; left uncited
Cited 3 spots in "text-to-cite" with 4 papers. No supporting paper was found for sentence 4; add keywords and try again.
After the references are found
English and Chinese literature together
Each spot gets two English query sets from different angles such as method, problem and application, plus two Chinese sets when the text is in Chinese.
Reference list updated too
Citations link to the original so you can check them; if there is a reference list, new papers are added in its existing format.
Keep writing in the workspace
The cited text stays in the workspace, where you can polish it, lower similarity, or keep writing from these papers.
Paper polishingAcknowledged in published SCI papers
Published papers state in their acknowledgments or AI-use statements that Acadwrite was used to polish the language.
- Theranostics2024Microfluidic chips in female reproduction: a systematic review of status, advances, and challenges
During the preparation of this work the authors used Acadwrite in order to improve readability, grammar and language style.
- Frontiers in Neuroscience2024Physical exercise for brain plasticity promotion an overview of the underlying oscillatory mechanism
We would like to thank Acadwrite (acadwrite.cn) for English language editing.
- Poultry Science2025Effects of induced molting on lipid accumulation in liver of aged laying hens
During the preparation of this work authors used (https://acadwrite.cn/) in order to improve language.
- Aging Medicine2025The Mechanism and Potential Therapeutic Strategies of Vascular Aging
Acadwrite AI (https://acadwrite.cn/apps/polish) was utilized for polishing the English language of the manuscript to enhance readability.
- Interactive Journal of Medical Research2025Evolution of Artificial Intelligence in Medical Education From 2000 to 2024: Bibliometric Analysis
During the preparation of this work, the authors used Acadwrite to improve readability, grammar, and language style.
AI citation finder FAQ
Didn't find your answer? Citation finding is charged by actual usage; see credit details. See credit details
Are the references real?
Every citation comes from a real academic database search and was checked against its abstract for whether it supports the sentence. Citations link to the original so you can verify them.
Will it pad citations to fill gaps?
No. Papers that are on topic but cannot support the specific claim are not used, and spots with no suitable paper are left uncited and listed in the report.
Will it change my text?
No. Citations go just before the end punctuation and no other word changes.
Which sentences get citations?
Sentences describing prior work and methods, citing others' data and conclusions, restating others' views or comparing with other methods. Your own contributions, common knowledge and transitions do not, and sentences already cited are skipped.
How many papers per spot?
1–3, chosen by how well each abstract supports the sentence.
Can it find Chinese references?
Yes. Choose Chinese as a search language; for Chinese text it also writes Chinese queries.
What about reference formatting?
Citations are inserted author–year and linked to the original; an existing reference list is updated in its format. For GB/T 7714 or other styles, ask for it in the workspace.
How much does it cost?
Sign-in is required. Citation finding runs in the workspace and is charged by tokens used, at 20 credits per million tokens. See the credits page for details.