TotalText — TotalText_RS_think

Rejection-sampled from the TotalText train split. This split holds the accepted items, with the model's reasoning trace.

rows	291
QA pairs	291
shards	4
accepted / rejected (whole family)	291 / 313
accept rate	48.2%
verifier	lines

How the data was produced

A VLM answers every question at temperature 0 with reasoning enabled. Its answer is compared with the official ground truth by the verifier described below; matches go to TotalText_RS_think and TotalText_RS_nothink, non-matches go to TotalText_rejected with the ground truth preserved.

The question field is byte-identical to the prompt the model actually received, including any appended instruction, so the stored training text and the inference-time input cannot drift apart.

How a match is decided

Every rule below is stated with its exact constants and a runnable reference implementation, so any accept/reject decision in this dataset can be recomputed. All snippets assume:

def levenshtein(a, b):
    if a == b: return 0
    if not a: return len(b)
    if not b: return len(a)
    prev = list(range(len(b) + 1))
    for i, ca in enumerate(a, 1):
        cur = [i] + [0] * len(b)
        for j, cb in enumerate(b, 1):
            cur[j] = min(prev[j] + 1, cur[j - 1] + 1, prev[j - 1] + (ca != cb))
        prev = cur
    return prev[len(b)]

lines — multi-instance transcription

When an image holds several independent text instances, the annotation and the model may split and order them differently, so the comparison is over the set of normalised alphanumeric tokens rather than over the string.

def token_set(x):
    out = set()
    for part in re.split(r"[;\n,]+", x or ""):
        for w in part.split():
            t = re.sub(r"[^a-z0-9]", "", w.lower())
            if t: out.add(t)
    return out

def lines_pair(pred, gt):
    return 1.0 if token_set(pred) == token_set(gt) else 0.0

Second pass: model-as-judge

String verifiers cannot judge free-form answers at all, and they under-credit answers that are correct but differently phrased. Every item the verifier rejected is therefore re-examined by the same VLM in a text-only pass that sees the question, the reference and the candidate — no image, no re-generation, it only grades text that was already produced.

The judge is instructed to compare against the reference and never to solve the question itself, so an answer is not promoted merely because the model believes it is true. It answers CORRECT, WRONG, or UNSURE; numbers must agree, and UNSURE is reserved for references that are themselves unusable or ambiguous. UNSURE is never counted as correct — only a CORRECT verdict moves a row into the accepted splits. WRONG and UNSURE both remain in the rejected split, so an unresolved case is never silently promoted to a pass.

Calibration: run over items the string verifier had already accepted, the judge agreed on 59/60. Spot-checking its overturns shows it discriminates rather than rubber-stamps — it keeps \frac{L}{n} rejected against a reference of \frac{L}{n-1}, and returns UNSURE when the model argues no finite answer exists against a numeric reference.

What answer holds

In the accepted splits answer is the model's own verified output — the text that passed the verifier — so it agrees with the reasoning in think. The original reference annotation is kept beside it as org_answer. Train on answer; use org_answer when you need the source label.

The two differ more often than one would guess: the verifier accepts on meaning, not on characters, so casing, spacing and formatting routinely differ (bridgestone vs Bridgestone, ( n x m ) vs (n x m)). Storing the reference in the answer slot would pair a model's reasoning with someone else's wording.

In the rejected split the roles are different: answer is the official ground truth and wrong_vlm is what the model said instead.

Schema

• image — HF Image() — renders directly in the dataset viewer
• qa — list of {question, answer, org_answer, think}
• clean_meta — per-row provenance and the cleaning policy that admitted it
• image_sha256 — content hash of the image

Examples

Three rows taken straight from this split. Images are the original files as stored in the dataset, and long fields are shown in full.

Example 1

field	value
question	Transcribe all legible text in this image. Use one line per text instance.
answer	SANIKA - PARK
org_answer	SANIKA PARK

think (112 chars)

The user requires the text from the image. I see a yellow archway with red text.
The text reads "SANIKA - PARK".

Example 2

field	value
question	Transcribe all legible text in this image. Use one line per text instance.
answer	TRACK 4 CROSSING
org_answer	TRACK CROSSING 4

think (93 chars)

The user requires the text from the sign. I will transcribe the words exactly as they appear.

Example 3

field	value
question	Transcribe all legible text in this image. Use one line per text instance.
answer	MALUKU
org_answer	MALUKU

think (79 chars)

The user wants the text from the sign. I see a white sign with black lettering.

Cleaning applied before rejection sampling

• Provenance check. Each QA item records where it came from; items whose recorded source does not resolve to the image they are attached to are dropped.
• Task purity. The pool's InfinityMM slice assigns samples to a family by image match rather than by dataset origin, which pulls unrelated general-VQA questions into benchmark families. Families where that content measured off-task keep only their genuine benchmark QA.
• Multiple-choice normalisation. Where options had been flattened into prose, they are re-rendered as a lettered block and the reference is rewritten to the bare letter, so the task is well posed and the answer is unambiguous.