見
00PROLOGUE研究を、漫画で読む。

What do you want
to understand?

Ask in plain words. Kenshō finds real research papers on the web, reads them, and draws them as manga chapters you can actually follow — every claim cited to the paper.

Kai: Wait… I just read a 40-page research paper?!

Prof. Hoshi: Every paper is a story. You only needed it drawn.

Sumi: I swim the web with TinyFish and fetch the real papers!

Prof. Hoshi: Every claim stays tied to the exact words in the paper.

第一話 · how it feels1
第一話 · EPISODE 1

How it works

Three steps from a question to a chapter you understand.

arxiv.orgopenreview.netncbi.nlm.nih.gov
01探す

Ask a question

TinyFish searches research indexes and open repositories — arXiv, OpenReview, PubMed Central and more — and brings back real papers with their original links.

02読む

Pick a paper

Each result says whether its full text is open or only the abstract is. Kenshō reads what's openly available and never gets around paywalls.

検証VERIFIED
03わかる

Read it as manga

Chapters with characters, diagrams and key terms. Analogies are marked as analogies; every research claim links to a verbatim quote from the paper.

第二話 · EPISODE 2

On the shelf

Manga already drawn from real papers on this Kenshō.

RECENTLY EXPLORED

第三話 · EPISODE 3

Start with a topic

Not sure what to ask? Open one of these.

第四話 · EPISODE 4

Grounded, not made up

Kenshō separates what the paper says from how we explain it. Research claims carry citation chips that open the exact sentence from the paper; analogies wear a 喩 ANALOGY tag.

Every quote is checked against the text that was actually retrieved. Matches earn a red 検証 stamp; anything that couldn’t be matched is shown as unverified. Chart numbers must appear in a verified quote, or the chart isn’t drawn.

論文より · FROM THE PAPERS2 · Abstract
“We propose a new simple network architecture, the Transformer, based solely on attention mechanisms…

In plain words: the authors replace recurrence entirely with attention.

検証VERIFIED
How a claim panel looks · quote from Vaswani et al., “Attention Is All You Need” (2017), arXiv:1706.03762