正在加载视频...

视频加载失败

Use this FREE tool to identify hallucinated & fake references. This tool is used by many scientific journals. 1. Go to and log in 2. Upload the paper or list of references 3. Reviewer3 will check each reference 4. It will display a YES or NO for each reference...

10,192 次观看 • 7 个月前 •via X (Twitter)

4 条评论

Juzly 的头像
Juzly7 个月前

@reviewer3com I used ChatGPT, but the references were fake.

Faheem Ullah 的头像
Faheem Ullah7 个月前

@reviewer3com This happens a lot. @reviewer3com can help to identify these fake references.

WriteForm 的头像
WriteForm7 个月前

@reviewer3com Amazing. This solves so many problems with people claiming so many fake citations from the hallucinations

Faheem Ullah 的头像
Faheem Ullah7 个月前

@reviewer3com Exactly. Glad you found it useful

相关视频

How to read a research paper 10x faster? PhD Students are often under time pressure. Many research papers but little time to read. Meet Bohrium – a tool that can 10x your paper reading. 𝐇𝐨𝐰 𝐜𝐚𝐧 𝐁𝐨𝐡𝐫𝐢𝐮𝐦 𝐡𝐞𝐥𝐩 𝐲𝐨𝐮 𝐫𝐞𝐚𝐝 𝐬𝐦𝐚𝐫𝐭𝐞𝐫? 1. Go to and log in. 2. Click on 𝑁𝑒𝑤 𝐶ℎ𝑎𝑡 and then on 𝐿𝑖𝑡𝑇𝑎𝑙𝑘 3. Upload the paper you want to read 4. Click on 𝑆𝑢𝑚𝑚𝑎𝑟𝑖𝑧𝑒 from the list. 5. Bohirum will summarize the paper for you. 6. This will help you understand the gist of the paper. 7. Now click on 𝐸𝑥𝑡𝑟𝑎𝑐𝑡 𝐹𝑖𝑛𝑑𝑖𝑛𝑔𝑠 8. Bohrium will extract key findings from the paper. 9. Go through each of the key finding. 10. Bohrium links each finding to respective part 11. This way you can see the links to the paper. 12. In the similar manner, you can also ↳ Extract information about research methods ↳ Academic concepts used in the paper 13. Bohrium can also help you in your writing 14. For example, you can use it to create diagrams 15. You can use Bohrium to create ✓ Graphical abstract ✓ Mechanism diagram ✓ Flowchart ✓ Concept cover 𝐖𝐡𝐚𝐭 𝐢𝐬 𝐬𝐩𝐞𝐜𝐢𝐚𝐥 𝐚𝐛𝐨𝐮𝐭 𝐁𝐨𝐡𝐫𝐢𝐮𝐦? → You can upload and chat with multiple papers together → Instead of only summarizing, it helps you in deep understanding → It links key findings to exact sections of the paper → It works like research-focused but free NotebookLLM Try it today:

Faheem Ullah

27,133 次观看 • 8 个月前

What started as building a personal taste.md skill for myself, turned into building a pipeline to create any taste as a skill. The most important piece is references. This is where you should spend time. If the references suck, so does the skill. I find that references cropped tightly on details in high resolution work the best. Each image gets analyzed by both Opus 4.7 and GPT 5.5. The analysis is based on why the reference is successful as a piece of design - not what it does functionally. Using two models helps rule out biases and gaps from each. The models focus on layout, spacing, typography, rhythm, composition, hierarchy, etc. At the end, each image has: reference-01/ - opus-4-7-analysis.md - gpt-5-5-analysis.md Then we fuse them together using GPT 5.5 - but the md files are anonymized so 5.5 doesn't prefer itself. reference-01/ - fused-analysis.md reference-02/ - fused-analysis.md etc. After fusion, we have one synthesized analysis per reference. Now the goal is to combine all of those into a single rule set. This is where chunking matters. If you ask one model to combine 100 image analyses at once, the result becomes too broad. It summarizes instead of preserving the granular design rules we want. Instead we chunk the fused analyses into smaller groups. Each group gets merged into a chunk-level synthesis, usually from around 6 to 8 image notes at a time. Then one final model pass fuses those chunks into a single md rule set. Finally, using the rule set, we write a skill of concrete instructions. It enforces constraints, uses imperative wording, and avoids vague taste words.

Jaytel

59,853 次观看 • 3 个月前