Practical guidance for literature discovery, source selection, PDF reading, evidence comparison, responsible AI-assisted drafting, and common problems.
⌕
15
answers
Help topic
Research workflow
From a focused question to a traceable research brief.
3
What is Byleron Three?
Byleron Three is an academic research workspace for discovering literature, selecting useful sources, comparing evidence, and developing a cited research brief. It supports research preparation and synthesis; the researcher remains responsible for the question, selections, interpretation, and final claims.
What to do
Write a focused research problem or question.
Choose a discovery mode and review the returned sources.
Keep, compare, and write from evidence you have checked.
Use Quick for a fast, bounded starting set; Semantic when the wording of a question needs broader conceptual matching; Literature Scan when you want a structured review of a larger source set; and Chat with Research when you want to ask follow-up questions about the sources already selected for the current research record.
What to do
Start with the smallest mode that can answer your immediate question.
Use Literature Scan when you need to compare a wider body of work.
Use Chat with Research only after reviewing the selected evidence base.
A research record keeps the question, search activity, selected sources, evidence, comparisons, and generated writing together. This makes it easier to return to a line of reasoning and inspect which materials informed a draft.
What to do
Give the record a recognisable research-focused name.
Review selected sources before relying on a generated synthesis.
Keep notes about inclusion, exclusion, and scope decisions.
Select readable, relevant material and preserve its context.
3
How should I select sources?
Select sources because they are relevant to the research question, credible for the purpose, and sufficiently described to evaluate—not merely because they agree with an expected conclusion. Keep material that complicates or limits an emerging account when it matters.
What to do
Check title, authorship, year, abstract, source type, and relevance.
Record why central, contrasting, or excluded sources matter.
Read important full texts before making high-stakes claims.
The PDF library stores research PDFs connected to your record and extracts readable text where possible so that it can be searched and discussed in context. Capacity is measured by documents, file size, and extracted text characters according to your plan.
What to do
Upload only material you have the right to use.
Check extracted text against the visible PDF before quoting it.
Use the original PDF when formatting, tables, or page layout affect meaning.
Some PDFs contain images rather than a usable text layer, use unusual fonts, have restricted access, or contain complex columns and tables. In these cases extraction can be incomplete or inaccurate even though the document looks correct on screen.
What to do
Open the original PDF and compare a short passage.
Try a clearer or publisher-provided version when available.
Do not rely on extraction alone for a quotation you intend to publish.
Move from selected sources to a reviewable, cited synthesis.
3
What does Literature Scan do?
Literature Scan organises a larger selected source set into a comparative research view. It can help identify recurring positions, differences, evidence gaps, and limitations, but it does not establish a systematic review or meta-analysis unless you design and document those methods separately.
What to do
Set a clear scope and eligibility logic before interpreting patterns.
Review which sources and passages support each comparison.
Separate reported findings from your own interpretation.
Content Creator helps turn a selected evidence base into a structured draft, such as an overview, literature section, or research-oriented explanation. It is a drafting aid: edit wording, verify every citation, and ensure the argument reflects your method and intended audience.
What to do
Select a purpose and writing mode.
Review the cited evidence before retaining a statement.
Revise the draft in your own scholarly voice and format.
Use citations to guide verification, then check the cited source itself. A citation does not automatically prove that a sentence accurately represents a study’s design, sample, result, or limitation. Ensure that wording, scope, and strength of claim match the original material.
What to do
Open the cited source or record before reusing the claim.
Check publication details in your reference manager or publisher record.
Qualify claims when evidence is indirect, limited, or inconsistent.
Understand access before beginning a larger research task.
3
What counts toward a monthly allowance?
Allowances are tracked separately for research actions such as Quick, Pro, Literature Scan, Chat with Research, Content Creator, and the PDF library. An allowance is used only when the related action begins successfully; your available plan determines the monthly maximum.
What to do
Check the allowance summary before starting a large task.
Use Quick to narrow a topic before running a larger scan.
Upgrade only when the current research workload genuinely requires it.
Your existing research records remain available. You can wait for the next monthly reset, narrow the task, or choose a plan with more capacity. Do not repeatedly submit the same request after a limit notice; that will not increase the allowance.
What to do
Save the research question and selected evidence.
Check the reset information in Account settings and usage.
Where can I manage my plan or request help with billing?
Membership pages manage plan selection and billing. Use Contact support for account access, cancellation, downgrade, refund, or institution enquiries. Explain the request without sending payment-card details in a message.
What to do
Open Membership Levels to compare available plans.
Use Contact support for account or billing concerns.
Keep your account email and request summary available.
Why did my search return few or unsuitable sources?
A question may be too broad, too narrow, highly local, newly emerging, or expressed with terms that differ from the literature. Start by clarifying the population, concept, setting, and outcome or phenomenon of interest, then try alternative scholarly terms.
What to do
Replace broad phrases with a specific research concept.
Try synonyms, spelling variants, and related disciplinary terms.
Use a broader discovery run before judging the available evidence.
Check the research scope, every citation, the wording of each claim, and the treatment of disagreement or uncertainty. AI features can help organise and draft, but they should not decide relevance, invent evidence, or make methodological claims for you.
What to do
Read the cited source passage for important claims.
Remove unsupported, overconfident, or context-free wording.
Add your method, rationale, and limitations before publication.
Describe the action, what you expected, what occurred, and any visible message. Include browser and device details, the research mode, and a screenshot only after removing confidential information. Support normally replies within 10 business days.
What to do
Copy the exact visible error when available.
State whether the issue affects one record or all records.
Remove participant data, credentials, and copyrighted files from attachments.
Try a shorter phrase, the name of the workspace area, a file type, or the exact words from the error message.
Support
Still need help?
Tell us what you were trying to research, the mode you selected, what you expected, and what happened instead. Include the exact error, browser, source type and approximate file size, and a screenshot only when it does not expose confidential research material.
Never send passwords, API keys, payment details, participant identifiers, or confidential source files in an ordinary email.