A vision-language model read each image together with its QA and judged the item. The pass is not a filter that only removes rows — it rewrites answers it finds wrong but salvageable, drops what it cannot salvage, and adds QA where the image supports more than the source provided. Each row carries the judge's own record in clean_meta, including the cleaner identity, the policy it applied, and its per-item scores for legibility, richness and coverage.

A large share of the QA here was written by the cleaning pass, not by the original dataset. Across the pool that share runs from roughly half to over 80% of a family's turns, and it is reported in the table above. Those rows have an empty org_answer because no pre-clean original exists. Treat them as model-generated supervision: they were judged against the image, but they are not human annotation, and model-written QA is where formatting defects are most likely.

The effect on the answers that were carried over is substantive rather than cosmetic. In this pool the pass corrects values, not just wording — a curve's critical point restated from 4.00 to 2.00, a computed ratio from 1 to Approximately 1.33 — and for some families it removed the content entirely, which is why those families are absent here rather than published empty.

`answer` vs `org_answer`

answer — the cleaned answer. Train on this.
org_answer — the pre-cleaning answer from the same (image, question) in the uncleaned pool. It is empty for QA the cleaning pass added, which have no original.

Where the two differ, the difference is the correction. Keeping both makes every edit auditable instead of asking you to take the cleaning on trust.

Schema

image — HF Image(), renders directly in the dataset viewer
image_sha256 — content hash of the image
n_qa — number of turns attached to this image
qa — list of {question, answer, org_answer}, in source order
family / clean_meta — provenance and the judge's per-row record

A row with more than one entry in qa is a conversation over a single image, not a set of independent questions — the turns depend on each other and the image belongs to all of them. Keep them together and in order when training.

Parquet row groups are written at ~32 MB so the dataset viewer pages without stalling.

What is NOT claimed

No rejection sampling was run against this family: there is no accept/reject verdict per row, and answer is the cleaned reference rather than a model generation that a string verifier confirmed. For that, use the *_RS_think / *_rejected datasets in elliot-mllm.

Examples

10 rows taken straight from this dataset. Images are the original files as stored here - no downscaling, no recompression - and every turn is shown in full.

Example 1

#	question	answer
1	Describe the visual style of the text and the background color.	White handwritten text on a black background.

Example 2 — 3 turns on one image

#	question	answer
1	Please write out the expression of the formula in the image using LaTeX format.	\frac { - b + \sqrt { b ^ { 2 } - 4 a c } } { 2 a }
2	What is the background color of the image?	Black
3	Is the formula in the image typed or handwritten?	Handwritten

Example 3 — 4 turns on one image

#	question	answer
1	Please write out the expression of the formula in the image using LaTeX format.	\sum _ { m = 0 } ^ { M } \sum _ { n = 0 } ^ { N } a _ { m , n }
2	What is the background color of the image?	Black
3	What is the writing style of the formula?	Handwritten
4	What color is the text in the image?	White

Example 4 — 2 turns on one image

#	question	answer
1	Please write out the expression of the formula in the image using LaTeX format.	f = d ^ { 5 } a _ { 4 } + 5 b d ^ { 4 } a _ { 5 } = - 3 2
2	What are the visual characteristics of the text in the image?	The text appears to be handwritten in white ink on a black background.

Example 5 — 2 turns on one image

#	question	answer
1	Please write out the expression of the formula in the image using LaTeX format.	\lim _ { x \rightarrow \infty } \int _ { 0 } ^ { x } e ^ { - y ^ { 2 } } d y = \frac { \sqrt { \pi } } { 2 }
2	Describe the visual style and color scheme of the equation in the image.	The equation is handwritten in white script against a solid black background.

Example 6 — 2 turns on one image

#	question	answer
1	Please write out the expression of the formula in the image using LaTeX format.	N
2	What are the colors of the character and the background in the image?	The character is white and the background is black.

Example 7 — 3 turns on one image

#	question	answer (cleaned)	org_answer (pre-judge)
1	Please write out the expression of the formula in the image using LaTeX format.	T^{C+L}(m)	T ^ { c + L } ( m )
2	What are the colors of the text and the background in the image?	The text is white and the background is black.	(unchanged)
3	How is the term 'C+L' positioned relative to the letter 'T'?	It is written as a superscript above and to the right of the letter T.	(unchanged)

Example 8 — 4 turns on one image

#	question	answer
1	Please write out the expression of the formula in the image using LaTeX format.	m ^ { b }
2	What is the background color of the image?	Black
3	What is the color of the mathematical expression?	White
4	How is the letter 'b' positioned relative to the letter 'm'?	It is positioned as a superscript to the upper right.

Example 9 — 2 turns on one image

#	question	answer
1	Please write out the expression of the formula in the image using LaTeX format.	\log _ { 2 } \frac { 1 } { 2 } + \log _ { 4 } \frac { 2 } { 4 }
2	Describe the visual style and color scheme of the text in the image.	The text is handwritten in white on a black background.

Example 10 — 2 turns on one image

#	question	answer
1	Please write out the expression of the formula in the image using LaTeX format.	\sum _ { m } f ( m + 3 )
2	What are the colors of the text and the background in the image?	The text is white and the background is black.

Downloads last month: 28

Total file size:

119 MB

You need to agree to share your contact information to access this dataset

chrome_writting_cleaned

How this was cleaned

answer vs org_answer