Download data/tasks/presentations-008/rubric.json from Accio-Lab/VisualArtifactSpec: direct link, hf CLI and curl.
- Browser
- Download file 178 kB
-
https://huggingface.co/datasets/Accio-Lab/VisualArtifactSpec/resolve/main/data/tasks/presentations-008/rubric.json
- Command line
-
hf download hf://datasets/Accio-Lab/VisualArtifactSpec/data/tasks/presentations-008/rubric.json
-
curl -L -o rubric.json https://huggingface.co/datasets/Accio-Lab/VisualArtifactSpec/resolve/main/data/tasks/presentations-008/rubric.json
178 kB
| [ | |
| { | |
| "id": "IF01", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the delivered PDF is exactly 8 pages, corresponding to compilable Beamer TEX content. Verification method: read the actual PDF page count, and check content correspondence by normally compiling the source document; do not substitute the literal count of frame environments in the TEX for the actual slide page count. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 3: create **eight Beamer TEX slides plus PDF**.", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G01" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF02", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the eight slides include a cover/title page. Verification method: check that a cover or title page exists among the 8 pages of content and is counted in the total page count. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 15: slide counts include the cover and conclusion..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G02" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF03", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the eight slides include a conclusion page. Verification method: check that a conclusion page exists among the 8 pages of content and is counted in the total page count. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 15: slide counts include the cover and conclusion..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G02" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF04", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the base configuration has 8 attention heads. Verification method: find the base row or equivalent statement and check that the number of heads is 8. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 9: The base has 8 heads and dev BLEU 25.8.", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G03" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF05", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the base configuration's dev BLEU is 25.8. Verification method: find the base row or equivalent statement and check that the BLEU value is 25.8. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 9: The base has 8 heads and dev BLEU 25.8.", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G03" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF06", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: 1 attention head corresponds to dev BLEU 24.9. Verification method: find the head-count-to-score correspondence table or chart and check that 1 head corresponds to 24.9. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 9: 1/4/16/32 heads correspond to 24.9/25.5/25.8/25.4..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G03" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF07", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: 4 attention heads correspond to dev BLEU 25.5. Verification method: find the head-count-to-score correspondence table or chart and check that 4 heads correspond to 25.5. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 9: 1/4/16/32 heads correspond to 24.9/25.5/25.8/25.4..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G03" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF08", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: 16 attention heads correspond to dev BLEU 25.8. Verification method: find the head-count-to-score correspondence table or chart and check that 16 heads correspond to 25.8. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 9: 1/4/16/32 heads correspond to 24.9/25.5/25.8/25.4..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G03" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF09", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: 32 attention heads correspond to dev BLEU 25.4. Verification method: find the head-count-to-score correspondence table or chart and check that 32 heads correspond to 25.4. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 9: 1/4/16/32 heads correspond to 24.9/25.5/25.8/25.4..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G03" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF10", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the metric is explicitly labeled as newstest2013 development-set BLEU, not test-set BLEU or accuracy. Verification method: check whether the places where the metric appears are labeled newstest2013 and development/dev, and exclude incorrect labels such as test BLEU, test set BLEU, or accuracy. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 10: Identify the metric as newstest2013 development-set BLEU, not test BLEU or accuracy..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G04" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF11", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: explain that in this experiment the dimension of each attention head also changes with the number of heads. Verification method: find the setup page or results note and confirm that there is an explanation that head dimensions change with the number of heads. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 11: Explain that head dimensions also change in this experiment.", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G05" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF12", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: do not claim that all other parameter values remain fixed. Verification method: search for all other parameters remain fixed or an equivalent statement, and confirm that this incorrect claim does not appear. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 11: do not claim all other parameter values remain fixed..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G06" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF13", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: do not introduce facts unsupported by the prompt and originals. Verification method: spot-check data, conclusions, captions, and textual explanations, and confirm that all are supported by the prompt, PDF, CSV, or visible explanations. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 5: Do not introduce unsupported facts..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G07" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF14", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: when the original PDF/webpage conflicts with auxiliary TXT/CSV files, use the original PDF/webpage content. Verification method: compare the relevant values or labels in the original PDF/webpage and the auxiliary TXT/CSV; if they conflict, check that the slides use the original source value. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 5: Original PDFs/webpages take precedence over conflicting auxiliary TXT/CSV files..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G07" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF15", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the content is based on the experimental setup and the Table 3 base/(A) head ablations. Verification method: check that the displayed ablation setup and data correspond to the Table 3 base/(A) head ablations, rather than other experiments or table entries. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Material basis (relative to inputs/): input/workspace/reading-scope.md original line 3: Use the experimental setup and Table 3 base/(A) head ablations..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G08" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF16", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the slides are organized around the ablation question, experimental setup, results, and conclusion, and connect to one another. Verification method: check page by page whether the four parts—ablation question, setup, results, and conclusion—are included and connected, and whether the narrative order and references are consistent. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 27: Connect the ablation question, setup, results, and conclusion without confusing variables..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G09" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF17", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the results presentation makes head counts and metrics directly comparable. Verification method: check whether the results page uses the same table, chart, or unified coordinates to display each head count and its BLEU, facilitating horizontal comparison. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 28: Make head counts and metrics comparable, with readable table headings, units, and conditions..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G10" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF18", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the headings of the results table/chart indicate head count and the BLEU metric. Verification method: check whether the headings/axis titles of the results table or chart explicitly indicate head count and the BLEU metric. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 28: Make head counts and metrics comparable, with readable table headings, units, and conditions..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G10" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF19", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: use the Metropolis Beamer theme style. Verification method: check whether the TEX preamble uses the metropolis theme or includes an equivalent theme source file. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 11: Retain the Metropolis style..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G05" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF20", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: retain the color system of the specified Metropolis theme. Verification method: check whether the slide color scheme follows the color definitions of the provided Metropolis theme. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Material basis (relative to inputs/): input/workspace/README.md original line 5: Retain the specified theme's colors, title, and footer system; replace example content..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G11" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF21", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: retain the title system of the specified Metropolis theme. Verification method: check whether page titles, frametitles, or title layouts follow the title style of the provided Metropolis theme. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Material basis (relative to inputs/): input/workspace/README.md original line 5: Retain the specified theme's colors, title, and footer system; replace example content..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G11" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF22", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: retain the footer system of the specified Metropolis theme. Verification method: check whether footer, author/institution/page number, and other elements follow the footer style of the provided Metropolis theme. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Material basis (relative to inputs/): input/workspace/README.md original line 5: Retain the specified theme's colors, title, and footer system; replace example content..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G11" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF23", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: replace the Metropolis example content and do not keep the template example text as body text. Verification method: search the template examples for content such as Hello world and A minimal example, and confirm that it has been replaced with the ablation content for this task. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Material basis (relative to inputs/): input/workspace/README.md original line 5: Retain the specified theme's colors, title, and footer system; replace example content..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G11" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF24", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: slide text does not overflow the content area. Verification method: check page by page whether text boxes, tables, and graphics exceed the page or content area and do not overlap one another. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 16: Avoid text overflow.", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G12" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF25", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: text, shapes, and data charts remain in editable form. Verification method: check whether the text, tables, shapes, or charts in the TEX/Beamer source exist as editable source/vector form rather than being rasterized as a whole. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 17: Preserve editable text, shapes, and data charts.", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G13" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF26", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: a full-slide raster image must not be the sole content of any page. Verification method: check page by page whether there is a slide consisting solely of a single PNG/JPEG or other raster image covering the whole page, with no editable text or other content. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 17: a full-slide raster image must not be the sole content..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G13" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF27", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the TEX source can be compiled to generate the corresponding PDF. Verification method: run the TeX compilation command in the preinstalled environment, and confirm that there are no missing packages or fatal errors and that the corresponding PDF is generated. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 18: supply compilable source, required assets, and PDF with TEX..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G14" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF28", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: provide the resources required for compilation together with the TEX. Verification method: check that the themes, fonts, images, and other resources referenced in the TEX are all provided in the delivery directory or a resolvable path. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 18: supply compilable source, required assets, and PDF with TEX..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G14" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF29", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: deliver a PDF file corresponding to the TEX. Verification method: check that an openable PDF exists in artifact/ and that its content corresponds to the TEX source. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 18: supply compilable source, required assets, and PDF with TEX..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G01" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF31", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: original materials remain read-only and unmodified [v7 verifiable boundary: check that the final input matches the frozen snapshot file by file; no proof is required that no write ever occurred during production. Public runtime inputs are guaranteed by read-only mounts.]. Verification method: check the available input integrity evidence against the content/hash of the frozen originals; public runtime read-only mounts can serve as a guarantee. Do not infer rewriting from file timestamps, missing history logs, or unretained intermediate processes. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 5 excerpt: Originals are read-only[…]artifact/ for deliverables..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G15" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF33", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: delivery.md explains the final file locations. Verification method: check that delivery.md lists the paths where the final files such as TEX, PDF, resources, and sources.md are located. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 19: explain final file locations and any necessary opening instructions in `delivery.md`.", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G16" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF34", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: if there are special opening or compilation steps, delivery.md gives the necessary instructions. Verification method: check whether delivery.md explains the compilation engine, how to open the PDF, or any non-default opening steps; if no special steps are needed, recording that there are no special requirements is sufficient. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 19: explain final file locations and any necessary opening instructions in `delivery.md`.", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G16" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF35", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: document the fonts used by the presentation. Verification method: check whether delivery.md or the artifact explains the fonts/font packages used. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Material basis (relative to inputs/): input/workspace/README.md original line 5: Available fonts may be configured for any retained source-language text; document fonts and compilation engine..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G11" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF36", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: document the compilation engine. Verification method: check whether delivery.md or the artifact explains the TeX compilation engine used. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Material basis (relative to inputs/): input/workspace/README.md original line 5: Available fonts may be configured for any retained source-language text; document fonts and compilation engine..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G11" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF37", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: cite the sources of facts and images in the artifact or sources.md. Verification method: check whether the slide footers/citation page or sources.md lists the sources of facts and images; if sources.md does not exist, the citations within the artifact must be complete. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 19: Cite facts and images in the artifact or `sources.md`..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G17" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF38", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: when product names appear, preserve them according to the originals. Verification method: extract the product/model names that appear in the slides and check them against the names in the PDF or prompt; they must not be rewritten or misspelled. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 3: Preserve product names, paper titles, and necessary terminology..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G18" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF39", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: preserve the titles of the cited papers. Verification method: check whether the cited paper titles match the PDF/webpage or the original prompt text, and have not been omitted or rewritten. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 3: Preserve product names, paper titles, and necessary terminology..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G18" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF40", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: preserve necessary terminology. Verification method: check whether necessary terms such as attention head, multi-head attention, BLEU, and newstest2013 are preserved with their original meaning. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 3: Preserve product names, paper titles, and necessary terminology..", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G18" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "IF41", | |
| "dimension": "Task Fulfillment / Compliance", | |
| "criterion": "Requirement: the slides are presented in English for researchers. Verification method: spot-check the language of titles, body text, table headings, captions, and conclusions on each page, and confirm that the main presentation is in English; product names, paper titles, and retained terminology may follow the originals. Pass criterion: fully meets this requirement; missing, incorrect, or partially met is No. Evaluation tool failure cannot be grounds for No. Verification boundary: verify against the frozen prompt and necessary originals/ground truth; the author's self-report does not replace product evidence. Task basis: prompt.md original line 3: in English for researchers at a group meeting.", | |
| "answers": { | |
| "Yes": "Complete all necessary checks for this criterion, with actual evidence supporting satisfaction of the requirement; equivalent implementations within the task's scope are allowed.", | |
| "No": "At least one necessary requirement is not satisfied; identify evidence of actual omissions, errors, or partial satisfaction." | |
| }, | |
| "compliance_groups": [ | |
| "G19" | |
| ], | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "C1", | |
| "dimension": "Content", | |
| "criterion": "Terminology and metric naming are accurate.\nMeasurement object: the BLEU metric in the slides is accurately stated as newstest2013 development-set BLEU, and the meanings of core terms such as head, head dimension, and dk are consistent with the usage in Table 3 of the paper; cross-page terminology drift belongs to O3 and is not evaluated again in this item.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG is not required.\nObservation basis: find where BLEU appears in the slides and confirm whether newstest2013 and development/dev are stated. Check whether incorrect substitutes such as test BLEU or accuracy appear. Check whether terms such as head, head dimension, and dk are consistent with the usage in Table 3 of the paper.\nScoring boundary: whether the terminology, concepts, and wording allow the audience of this task to understand accurately; IF checks factual implementation, and C1 does not add points for the number of terms.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most easily confused concepts and qualifications remain concise and accurate, and there is no meaning drift among summary, details, and actual examples.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.", | |
| "answers": { | |
| "-3": "Severe failure: the core understanding or usage goal of this item cannot be completed. Specific manifestation: calling the metric test-set BLEU or accuracy, or systematically misusing core terms such as head/dk. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently impaired. Specific manifestation: repeated dev/test confusion or writing BLEU as accuracy, affecting audience understanding. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: one imprecise metric naming, such as omitting newstest2013 or dev, but the context can still correct it. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "0": "Basically acceptable: the metric is accurately labeled as newstest2013 development-set BLEU, the terminology for head count and per-head dimension is consistent, and there is no substantive mixing; the necessary requirements of this item are fully met, with no substantive defects, and performance is ordinary.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: newstest2013 development-set BLEU is consistently stated in table headings, captions, and body text, and head count is distinguished from per-head dimension. Quality improvement must be demonstrated as: easily confused terms, quantities, or references receive brief clarification where they appear, reducing the need for guesswork. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: terminology is mutually consistent across titles, table headings, captions, and conclusions, and brief definitions help the audience distinguish dev BLEU from test BLEU. Quality improvement must be demonstrated as: definitions, qualifications, and actual labels support one another, and complex meanings do not require stitching across sections. the terminology system reaches the level of mature academic presentation, and metrics, conditions, and variable names are unambiguous everywhere and precisely correspond to the context of Table 3 of the paper. The most easily confused concepts and qualifications remain concise and accurate, and there is no meaning drift among summary, details, and actual examples. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages. Task-specific object: the BLEU metric in the slides is accurately stated as newstest2013 development-set BLEU, and the meanings of core terms such as head, head dimension, and dk are consistent with the usage in Table 3 of the paper; cross-page terminology drift belongs to O3 and is not evaluated again in this item. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most easily confused concepts and qualifications remain concise and accurate, and there is no meaning drift among summary, details, and actual examples. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. An unusually economical conceptual formulation makes a genuinely confusable distinction directly understandable, with less interpretation than an equally accurate, concise conventional explanation. Identify the distinction and the wording/representation that produces this extra understanding; merely correct terminology or fluent prose is not exceptional. Apply the exceptional benefit specifically to: the BLEU metric in the slides is accurately stated as newstest2013 development-set BLEU, and the meanings of core terms such as head, head dimension, and dk are consistent with the usage in Table 3 of the paper; cross-page terminology drift belongs to O3 and is not evaluated again in this item. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most easily confused concepts and qualifications remain concise and accurate, and there is no meaning drift among summary, details, and actual examples. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 9, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "C2", | |
| "dimension": "Content", | |
| "criterion": "Ablation data and experimental relationships are faithful.\nMeasurement object: correctly present base 8 heads dev BLEU 25.8 and 1/4/16/32 heads corresponding to 24.9/25.5/25.8/25.4, and explain that when the number of heads changes, the per-head dimension also changes; do not claim that all other parameter values remain fixed.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG is not required.\nObservation basis: check whether the 8-head base is 25.8 and whether 1/4/16/32 heads are 24.9/25.5/25.8/25.4. Find whether it explains that head dimensions also change with the number of heads. Check whether there is an 'all other parameters remain fixed' or equivalent incorrect statement. Check whether it lists or implies the corresponding dk values or h×dk=512.\nScoring boundary: the support relationship between task claims and evidence/data/actual process; does not re-evaluate graphic encoding or whether source documents are easy to find.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most complex key comparisons or state chains can still be checked item by item; approximations, limitations, and counterexamples are not concealed by the conclusion.\nScenario reference · content quality (no additional threshold): base has 8 heads, dev BLEU 25.8; 1/4/16/32 heads correspond to 24.9/25.5/25.8/25.4; the metric is newstest2013 development-set BLEU; head dimension changes with the number of heads (h×dk=512), and it must not be claimed that all other parameters are fixed.\nApplicable boundary for this task (applies to all relevant items, no additional scoring items): it is not required to present all Table 3 groups of the paper (e.g., (B)-(E)).\nApplicable boundary for this task (applies to all relevant items, no additional scoring items): preserve product names, paper titles, and necessary terminology, such as Transformer, multi-head attention, BLEU, and newstest2013.\nApplicable boundary for this task (applies to all relevant items, no additional scoring items): do not introduce facts unsupported by the prompt and originals.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.", | |
| "answers": { | |
| "-3": "Severe failure: the core understanding or usage goal of this item cannot be completed. Specific manifestation: core numerical errors or claiming that all other parameters are fixed, causing the ablation conclusion to be completely distorted. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently impaired. Specific manifestation: multiple head-count-to-score correspondence errors, or explicitly interpreting the experiment as changing only the number of heads. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: one inaccurate numerical or conditional statement, such as omitting a certain head-count condition, but the main relationship is still discernible. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "0": "Basically acceptable: the five head-count/score correspondences in the prompt are correct, and it explains that head dimensions change with the number of heads without claiming that all other parameters are fixed; the necessary requirements of this item are fully met, with no substantive defects, and performance is ordinary.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: the data are complete and clearly state that changes in the number of heads are accompanied by changes in per-head dimension, without claiming that all other parameters are fixed. Quality improvement must be demonstrated as: key claims correspond closely to the values, conditions, or process evidence used, making verification more direct. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: presents head count, dk, and dev BLEU side by side in a table or note, so that variable boundaries can be verified directly next to the data. Quality improvement must be demonstrated as: evidence selection, comparison basis, and conclusion strength work together to make clear what can and cannot be inferred. the data, variable linkage, and conclusion boundaries form a complete evidence chain, explicitly stating the relationship h×dk=512, without exaggerating the ablation interpretation. The most complex key comparisons or state chains can still be checked item by item; approximations, limitations, and counterexamples are not concealed by the conclusion. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages. Task-specific object: correctly present base 8 heads dev BLEU 25.8 and 1/4/16/32 heads corresponding to 24.9/25.5/25.8/25.4, and explain that when the number of heads changes, the per-head dimension also changes; do not claim that all other parameter values remain fixed. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most complex key comparisons or state chains can still be checked item by item; approximations, limitations, and counterexamples are not concealed by the conclusion. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. A non-obvious presentation of the evidence exposes an important conditional relationship that a flawless fact-and-citation account would leave the reader to reconstruct. Show the additional understanding and its support, without inventing a claim, requiring more data, or rewarding the count of sources. Apply the exceptional benefit specifically to: correctly present base 8 heads dev BLEU 25.8 and 1/4/16/32 heads corresponding to 24.9/25.5/25.8/25.4, and explain that when the number of heads changes, the per-head dimension also changes; do not claim that all other parameter values remain fixed. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most complex key comparisons or state chains can still be checked item by item; approximations, limitations, and counterexamples are not concealed by the conclusion. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 9, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "C3", | |
| "dimension": "Content", | |
| "criterion": "Information selection and tone suited to a group meeting.\nMeasurement object: select information according to the purpose of an academic group meeting briefing, highlight the ablation question, setup, results, and conclusion, keep the tone restrained and appropriately qualify the conclusion, and do not pile up irrelevant paper details.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG is not required.\nObservation basis: check whether there are large background sections or other experiments from the paper unrelated to the head-count ablation. Check whether the conclusion page presents limited ablation results as a general law. Check whether the overall tone is suitable for a research group meeting rather than a product launch.\nScoring boundary: content selection, density, tone, and necessary qualifications around the scenario purpose; not font size or page visual hierarchy.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; appropriate selection is still made at the most information-dense points, key points and exceptions are understandable, and there is no redundant expansion or oversimplification.\nTask scenario: an academic group meeting projected presentation for researchers in the relevant field, requiring English Beamer slides to clearly present the head-count ablation question, setup, results, and conclusion.\nScenario reference · audience (no additional threshold): researchers in the Transformer/NLP direction, familiar with terms such as BLEU and multi-head attention, able to read at group-meeting projection distance.\nScenario reference · use (no additional threshold): oral presentation at a group meeting accompanied by slides, helping the audience understand the experimental conditions, results comparison, and conclusion boundaries of the head-count ablation.\nScenario reference · content selection (no additional threshold): prioritize accurate variable explanation: changes in head count are accompanied by changes in per-head dimension; the metric is newstest2013 development-set BLEU; data and conditions facilitate horizontal comparison.\nApplicable boundary for this task (applies to all relevant items, no additional scoring items): a PPTX version is not required.\nApplicable boundary for this task (applies to all relevant items, no additional scoring items): non-prompt purchase entry points, brand marks, or extra cover design are not required.\nApplicable boundary for this task (applies to all relevant items, no additional scoring items): do not expand 'retain the Metropolis style' into a pixel-by-pixel reproduction of input/workspace/templates/metropolis/demo/demo.tex or a fixed single color value.\nApplicable boundary for this task (applies to all relevant items, no additional scoring items): Chinese output or Chinese-English bilingual output is not required.\nApplicable boundary for this task (applies to all relevant items, no additional scoring items): the TEX+PDF branch does not automatically require page-by-page PNG; the prompt explicitly requires PNG previews only for the PPTX branch.\nApplicable boundary for this task (applies to all relevant items, no additional scoring items): original PDFs/webpages take precedence over auxiliary TXT/CSV; in case of conflict, the PDF and webpage prevail.\nContent focus: head count and newstest2013 development-set BLEU, head dimensions changing with head count\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.", | |
| "answers": { | |
| "-3": "Severe failure: the core understanding or usage goal of this item cannot be completed. Specific manifestation: the content is irrelevant to the group meeting purpose or the conclusion is over-extrapolated, so the audience cannot grasp the main ablation thread. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently impaired. Specific manifestation: a large amount of irrelevant detail crowds out the core, or the tone is like marketing rather than an academic report. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: there is some off-topic content or insufficient qualification of the conclusion, but the main thread is still identifiable. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "0": "Basically acceptable: the selection centers on the ablation topic, the explanation fully retains the necessary conditions, and no unsupported generalization is made; suitable for a group meeting; the necessary requirements of this item are fully met, with no substantive defects, and performance is ordinary.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: the selection focuses on question-setup-results-conclusion, and the conclusion wording is appropriately qualified. Quality improvement must be demonstrated as: removing restatements that do not aid understanding or placing necessary qualifications where needed, making key points more direct. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: the information density of each page matches the presentation pace, and observed results are actively distinguished from causal explanations. Quality improvement must be demonstrated as: emphasis, depth of explanation, and tone suit the audience; streamlined content still retains necessary conditions. the selection of a mature academic briefing: necessary background and ablation evidence are treated with appropriate detail, and conclusion boundaries are clear and do not exceed what the data support. Appropriate selection is still made at the most information-dense points, key points and exceptions are understandable, and there is no redundant expansion or oversimplification. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages. Task-specific object: select information according to the purpose of an academic group meeting briefing, highlight the ablation question, setup, results, and conclusion, keep the tone restrained and appropriately qualify the conclusion, and do not pile up irrelevant paper details. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; appropriate selection is still made at the most information-dense points, key points and exceptions are understandable, and there is no redundant expansion or oversimplification. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. An unusually incisive selection resolves the tension between necessary qualifications and the audience’s main purpose: it conveys more useful understanding with no loss of required substance. Explain why a conventionally fully refined summary would still impose more interpretation; brevity, completeness or promotional impact alone is insufficient. Apply the exceptional benefit specifically to: select information according to the purpose of an academic group meeting briefing, highlight the ablation question, setup, results, and conclusion, keep the tone restrained and appropriately qualify the conclusion, and do not pile up irrelevant paper details. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; appropriate selection is still made at the most information-dense points, key points and exceptions are understandable, and there is no redundant expansion or oversimplification. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 7, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "O1", | |
| "dimension": "Organization", | |
| "criterion": "Argument path from question to conclusion.\nMeasurement object: the eight pages as a whole form a clear logical path from the ablation question, experimental setup, and results comparison to the conclusion, and the page order allows the audience to follow without jumping back.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG is not required.\nObservation basis: list the topic of each page by page number, and check whether it covers the question, setup, results, and conclusion. Check whether the head-dimension linkage is explained before the results page. Check whether the conclusion page responds to the opening ablation question.\nScoring boundary: the logic of the overall argument/decision/operation/concept path, not scored against a fixed structural template.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most complex branches or comparisons can still be incorporated into the main thread, and the closure is grounded with no dangling steps or repeated detours.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.", | |
| "answers": { | |
| "-3": "Severe failure: the core understanding or usage goal of this item cannot be completed. Specific manifestation: the page order is chaotic, and question, setup, results, and conclusion cannot form a followable path. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently impaired. Specific manifestation: key links are missing or the order is clearly reversed, requiring the audience to piece together the argument themselves. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: there is a main path, but a connection jumps at one point, affecting understanding. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "0": "Basically acceptable: it unfolds as question-setup-results-conclusion, and the path is followable but fairly conventional; the necessary requirements of this item are fully met, with no substantive defects, and performance is ordinary.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: each page has a clear connection, and key transition pages point out how the previous page's conclusion leads to the next page. Quality improvement must be demonstrated as: one key transition clearly connects the premise to the next step, making the main thread easier to follow. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: the path explicitly distinguishes changes in the independent variable from results comparison, and the conclusion page returns to the opening question. Quality improvement must be demonstrated as: the premises, development, and conclusions of each part refer back to one another, and order and length serve the explanation or decision of this task. the argument path is compact and has a presentation rhythm, with question, variables, evidence, and qualified conclusion progressing layer by layer. The most complex branches or comparisons can still be incorporated into the main thread, and the closure is grounded with no dangling steps or repeated detours. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages. Task-specific object: the eight pages as a whole form a clear logical path from the ablation question, experimental setup, and results comparison to the conclusion, and the page order allows the audience to follow without jumping back. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most complex branches or comparisons can still be incorporated into the main thread, and the closure is grounded with no dangling steps or repeated detours. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. A non-obvious organization of the existing argument or user path makes separate necessary parts illuminate one another and materially reduces reasoning or navigation steps. Compare the benefit with an equally complete, logically sound conventional order; no additional section or prescribed order is required. Apply the exceptional benefit specifically to: the eight pages as a whole form a clear logical path from the ablation question, experimental setup, and results comparison to the conclusion, and the page order allows the audience to follow without jumping back. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most complex branches or comparisons can still be incorporated into the main thread, and the closure is grounded with no dangling steps or repeated detours. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 6, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "O2", | |
| "dimension": "Organization", | |
| "criterion": "Local semantic relationships and table unfolding.\nMeasurement object: the semantic relationships among the title, table headings, captions, notes, and body text within a single page are clear, and the conditions for comparing head counts and metrics can be understood locally.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG is not required.\nObservation basis: check whether the results table has clear headings, units, and experimental conditions. Check whether captions or notes explain abbreviations and variables. Check whether the title and body text on a single page point to the same comparison.\nScoring boundary: the semantic order and relationships of explanations within one page, stage, node, or comparison unit.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most complex unit can also be understood in the required order, and local explanations are concise, self-contained, and consistent with the overall main thread.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.", | |
| "answers": { | |
| "-3": "Severe failure: the core understanding or usage goal of this item cannot be completed. Specific manifestation: table headings, units, or conditions within a single page are missing or contradictory, making the local part unreadable. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently impaired. Specific manifestation: the results page lacks necessary headings or condition explanations, and readers cannot determine the meaning of the numbers. Judge by the above actual impact; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: in a local unit, the relationship among objects, explanations, and results is unclear, and guessing or filling in is required to complete the necessary understanding. Provide evidence of actual impact in the relevant parts of this task concerning local semantic relationships and table unfolding; ordinary performance, harmless minor differences, or purely personal preference do not trigger a negative level.", | |
| "0": "Basically acceptable: the local information on each page is self-evident, and headings and units are readable; the necessary requirements of this item are fully met, with no substantive defects, and performance is ordinary.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: table headings, units, condition notes, and body text support one another, and local semantics are closed. Quality improvement must be demonstrated as: the arrangement of objects, explanations, and results in a local unit reduces one necessary re-check. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: uses grouping, annotation, or brief guiding sentences to explain the relationship among head count, dk, and BLEU on a single page. Quality improvement must be demonstrated as: the charts, text, conditions, and conclusions of each major unit form a locally understandable relationship. local unfolding reaches a professional presentation level, and complex comparisons are broken into readable layers without sacrificing accuracy. The most complex unit can also be understood in the required order, and local explanations are concise, self-contained, and consistent with the overall main thread. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages. Task-specific object: the semantic relationships among the title, table headings, captions, notes, and body text within a single page are clear, and the conditions for comparing head counts and metrics can be understood locally. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most complex unit can also be understood in the required order, and local explanations are concise, self-contained, and consistent with the overall main thread. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. An economical local organization makes a difficult relationship among conditions, evidence and objects understandable as one unit rather than assembled mentally from carefully grouped parts. Identify the semantic reconstruction it eliminates; proximity and tidy grouping alone do not qualify. Apply the exceptional benefit specifically to: the semantic relationships among the title, table headings, captions, notes, and body text within a single page are clear, and the conditions for comparing head counts and metrics can be understood locally. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition notes, and the highest-density formula/table pages only within the applicable scope of this item; the most complex unit can also be understood in the required order, and local explanations are concise, self-contained, and consistent with the overall main thread. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 5, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "O3", | |
| "dimension": "Organization", | |
| "criterion": "Cross-page identity and semantic inheritance.\nMeasured object: consistent title hierarchy, terminology, color semantics, and page number/footer identity across the eight pages; variable names remain consistent between the results page and the settings page.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: spot-check whether the title style, footer and page number are consistent on each page. Check whether terms such as head count, dk, and dev BLEU have the same name across pages. Check whether the same color always represents the same data series or emphasis meaning.\nRating boundary: semantic inheritance across pages/states, or closure of a single-page branch path with identity/action; do not force a newly added ending.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; complex branches, final states or resets still maintain identity and causal continuity, and the reader does not need later information to patch the current meaning.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.\nEvaluate meaning carried across the actual slides, not invented software branches or reset states.", | |
| "answers": { | |
| "-3": "Severe failure: the core understanding or usage goal of this item cannot be completed. Specific manifestation: cross-page terminology or color semantics change repeatedly, and the audience cannot confirm the same object. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: the meaning of key variable names or metric names drifts across pages, affecting comparison. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: a small amount of terminology or style inconsistency, but it can be corrected by oneself. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "0": "Basically qualified: cross-page style and terminology are unified; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: title hierarchy, footer, color semantics and variable naming are stable across pages. Quality improvement must be reflected as: one cross-page, cross-area or cross-state continuation explicitly preserves object identity and existing meaning. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: the terms and colors defined on the settings page are consistently reused on the results page and conclusion page and consciously echoed. Quality improvement must be reflected as: the main continuation, branches and ending form continuous semantics; the positions of repeated objects and necessary exceptions are traceable. cross-page identity is mature and unified, and the Metropolis title/footer system and content semantics form a stable presentation thread. Complex branches, final states or resets still maintain identity and causal continuity, and the reader does not need later information to patch the current meaning. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page. Task-specific object: Cross-page identity and semantic inheritance. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; complex branches, final states or resets still maintain identity and causal continuity, and the reader does not need later information to patch the current meaning. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. A purposeful carry-over lets the audience preserve and develop the right meaning through the task’s actual transitions, materially reducing reconstruction beyond fully refined referential consistency. Identify what understanding accumulates; do not invent branches, endings or reset states absent from the task. Apply the exceptional benefit specifically to: Cross-page identity and semantic inheritance. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; complex branches, final states or resets still maintain identity and causal continuity, and the reader does not need later information to patch the current meaning. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 4, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "V1", | |
| "dimension": "Visuals", | |
| "criterion": "Projection hierarchy and focus.\nMeasured object: each page has a clear visual focus at projection distance; titles, key numbers or main figures take priority over auxiliary explanations, and the audience can quickly locate the current presentation object.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: zoom to group-meeting projection scale and view the first visual landing point of each page. Check whether key numbers on the results page are more prominent than auxiliary text. Check whether decorative elements steal attention from content.\nRating boundary: the salience relationship among title, body, evidence and auxiliary information; evaluate the attention entry point, core focus and emphasis strength.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the densest area or state also shows the main information first and the basis and qualifications second, with no substantive attention competition between the whole and the parts.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.\nJudge negative visual grades by substantive harm and extent within this criterion, not a prerequisite of functional unusability. Harmless style differences are not defects; retain the original concrete defect examples.", | |
| "answers": { | |
| "-3": "Severe failure: This criterion’s visual goal is severely impaired. Specific manifestation: the page has no focus, all elements have equal weight, and it is impossible to tell where to look first during projection. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: the focus is drowned by decoration or secondary information, and key results are not prominent. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: a focus exists but is weak, requiring effort to find. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "0": "Basically qualified: each page has a focus, and the title and body hierarchy are distinguishable; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: key numbers or main figures form a stable focus through scale, position or whitespace. Quality improvement must be reflected as: a set of position, scale or contrast choices makes the main focus clearer, and auxiliary information does not compete for reading. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: the hierarchy changes with the presentation purpose: the settings page highlights conditions, the results page highlights comparison, and the conclusion page highlights qualifications. Quality improvement must be reflected as: the strength relationship among the main focus, evidence and necessary qualifications is coordinated, and reasonably adjusted with content complexity. the focus design is mature and restrained, and under projection the first glance lands on the evidence required by the current argument. The densest area or state also shows the main information first and the basis and qualifications second, with no substantive attention competition between the whole and the parts. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page. Task-specific object: Projection hierarchy and focus. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the densest area or state also shows the main information first and the basis and qualifications second, with no substantive attention competition between the whole and the parts. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. A non-obvious hierarchy brings the decisive relationship and its necessary qualification into attention together without competing emphasis. Demonstrate a first-look interpretation benefit beyond an impeccably clear primary/secondary hierarchy, not just stronger contrast or larger type. Apply the exceptional benefit specifically to: Projection hierarchy and focus. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the densest area or state also shows the main information first and the basis and qualifications second, with no substantive attention competition between the whole and the parts. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 5, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "V2", | |
| "dimension": "Visuals", | |
| "criterion": "Block proportions and whitespace.\nMeasured object: page block proportions, alignment and whitespace are suitable for 16:9 projection; tables and figures are not crowded; the relationship between page margins and the content area is stable.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: check whether tables are close to page margins or overflow the content area. Check whether the content-area proportions of the title page, results page and conclusion page are reasonable. Check whether alignment lines are stable across pages.\nRating boundary: area proportions among information blocks, grid, edges, center of gravity, density and whitespace; evaluate overall composition and spatial organization.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the area with the greatest density or state change still has a stable spatial order, and overall balance and local spacing both hold.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.\nJudge negative visual grades by substantive harm and extent within this criterion, not a prerequisite of functional unusability. Harmless style differences are not defects; retain the original concrete defect examples.", | |
| "answers": { | |
| "-3": "Severe failure: This criterion’s visual goal is severely impaired. Specific manifestation: blocks are severely unbalanced or content overflows, destroying necessary reading units. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: many pages are crowded or have large meaningless blank areas, affecting reading. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: local alignment or whitespace is inappropriate, but the whole is usable. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "0": "Basically qualified: the layout is balanced, whitespace is ordinary, and there is no substantive obstruction; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: the alignment of main blocks is stable, and there is enough breathing space around tables and figures. Quality improvement must be reflected as: block edges, scale or whitespace have a clear purpose, improving the spatial relationship of one class of content. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: different page types (title, settings, results, conclusion) have rhythmic proportional variation and all serve the content. Quality improvement must be reflected as: area proportions, alignment, center of gravity and whitespace jointly adapt to actual content, rather than mechanically applying the same layout. block proportions and whitespace reach a mature typographic level, and dense data pages still appear composed. The area with the greatest density or state change still has a stable spatial order, and overall balance and local spacing both hold. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page. Task-specific object: Block proportions and whitespace. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the area with the greatest density or state change still has a stable spatial order, and overall balance and local spacing both hold. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. A content-specific spatial solution reconciles competing demands of overview, detail and available space with exceptional clarity and balance. Identify the real space conflict resolved beyond a flawlessly fitted conventional composition; extra whitespace, density or a fashionable layout is not itself a benefit. Apply the exceptional benefit specifically to: Block proportions and whitespace. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the area with the greatest density or state change still has a stable spatial order, and overall balance and local spacing both hold. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 8, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "V3", | |
| "dimension": "Visuals", | |
| "criterion": "Type size, line spacing and projection readability.\nMeasured object: the type size and line spacing of body text, table headers, figure captions and footnotes are readable at group-meeting projection distance, and line breaks do not damage terminology and numbers.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: view the font sizes of body text, table headers and figure captions at projection size. Check whether long terms such as newstest2013 development-set BLEU are broken improperly. Check whether numbers and units are split apart.\nRating boundary: the complete visibility, type size, contrast, line spacing and line breaking of text in the actually delivered image; evaluate text presentation and reading at the target scale.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the longest necessary labels, small qualifications or complex symbols remain complete and easy to read, and the typographic quality of each area remains consistently stable.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.\nJudge negative visual grades by substantive harm and extent within this criterion, not a prerequisite of functional unusability. Harmless style differences are not defects; retain the original concrete defect examples.", | |
| "answers": { | |
| "-3": "Severe failure: This criterion’s visual goal is severely impaired. Specific manifestation: body text or table headers are too small to read under projection, and core information cannot be obtained. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: many pages have too-small type or too-tight line spacing, clearly affecting reading. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: one piece of necessary text is difficult to read completely at the scale specified by the prompt, or line breaking/contrast damages meaning acquisition. Provide actual impact evidence in the relevant part of this item's type size, line spacing and projection readability; ordinary performance, harmless minor differences or pure personal preference do not trigger a negative level.", | |
| "0": "Basically qualified: necessary body text, formulas, table headers, metrics and sources are readable at projection scale, and line breaking and type size have no substantive defects; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: the type hierarchy of body text and table headers is clear, and key numbers are not split by line breaks. Quality improvement must be reflected as: the type size, spacing or line breaking of one class of necessary text reduces reading burden, and cannot be based merely on being readable. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: the type-size system is coordinated with the Metropolis typographic tone, and long terms and numbers are stably readable on each page. Quality improvement must be reflected as: type size, line spacing, contrast and semantic line breaking are coordinated, and dense content remains comfortably readable at the scale specified by the prompt. mature typography: comfortable reading at projection scale, with hierarchy, line spacing and line breaking all carefully controlled. The longest necessary labels, small qualifications or complex symbols remain complete and easy to read, and the typographic quality of each area remains consistently stable. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page. Task-specific object: Type size, line spacing and projection readability. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the longest necessary labels, small qualifications or complex symbols remain complete and easy to read, and the typographic quality of each area remains consistently stable. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. Typographic structure makes a difficult meaning or relationship apparent as it is read, reducing parsing beyond fully refined legibility, spacing and hierarchy. Identify the specific semantic contribution at the required viewing scale; a distinctive font or decorative lettering is not sufficient. Apply the exceptional benefit specifically to: Type size, line spacing and projection readability. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the longest necessary labels, small qualifications or complex symbols remain complete and easy to read, and the typographic quality of each area remains consistently stable. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 7, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "V4", | |
| "dimension": "Visuals", | |
| "criterion": "Accurate decoding of charts and symbols.\nMeasured object: the meanings of axes, legends, units and arrows in tables, bar charts or schematic diagrams are accurate, and the correspondence between head count and BLEU is not misled by the graphics.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: check whether tables or charts indicate BLEU and the newstest2013 condition. Check whether the legend distinguishes head count from per-head dimension. Check whether arrows or annotations point to the correct object.\nRating boundary: the recognizability and semantic correspondence of graphic marks, lines, arrows, symbols and dynamic states; evaluate accurate decoding and reliable comparison.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the most complex connections, states or scale relationships remain reliably decodable, and the legend and actual content have no ambiguity or misleading.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.\nJudge negative visual grades by substantive harm and extent within this criterion, not a prerequisite of functional unusability. Harmless style differences are not defects; retain the original concrete defect examples.", | |
| "answers": { | |
| "-3": "Severe failure: This criterion’s visual goal is severely impaired. Specific manifestation: charts mislead about core data relationships, such as inverted axes or missing legends leading to misinterpretation. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: chart units or legends are missing, and readers easily misread the head-count–score correspondence. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: one necessary graphic, annotation or directional relationship cannot be accurately decoded, causing local ambiguity. Provide actual impact evidence in the relevant part of this item's accurate decoding of charts and symbols; ordinary performance, harmless minor differences or pure personal preference do not trigger a negative level.", | |
| "0": "Basically qualified: charts are correct and readable, and units and legends are complete; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: table headers, units, legends and value annotations are complete, and the correspondence between head count and BLEU is clear. Quality improvement must be reflected as: a set of marks, connections or comparisons makes object relationships more directly decodable, reducing guesswork about the chart. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: graphic encoding and tables complement each other, and readers can simultaneously compare absolute values and change trends without confusing variables. Quality improvement must be reflected as: graphic encoding, annotations and actual relationships support each other; complex comparisons do not require extra explanation to correct the chart's meaning. charts directly bind the relationship among head count, per-head dimension and BLEU through encoding or annotation, and readers can correctly interpret the variable boundaries without looking back at the body text. The most complex connections, states or scale relationships remain reliably decodable, and the legend and actual content have no ambiguity or misleading. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page. Task-specific object: Accurate decoding of charts and symbols. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the most complex connections, states or scale relationships remain reliably decodable, and the legend and actual content have no ambiguity or misleading. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. A visual encoding makes a difficult relationship directly perceptible instead of requiring mental reconstruction from fully correct labels and marks. Show what becomes easier to understand and verify fidelity to the source or required state; novelty and additional diagrams are not requirements. Apply the exceptional benefit specifically to: Accurate decoding of charts and symbols. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the most complex connections, states or scale relationships remain reliably decodable, and the legend and actual content have no ambiguity or misleading. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 5, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "V5", | |
| "dimension": "Visuals", | |
| "criterion": "Metropolis academic aesthetic integration.\nMeasured object: non-color-scheme graphic abstraction, typographic tone, images and component language are integrated into a Metropolis-style academic brief, without relying on piled-up decoration.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: check whether an irrelevant colorful template or heavy decoration is applied. Check whether tables, images and the title system belong to the same visual language. Check whether graphic abstraction serves data rather than decoration.\nRating boundary: audience, purpose, medium temperament, and the integration of graphic abstraction, typography, images and component surface language; pure color-scheme aesthetics is handed over to V9.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; even the hardest-to-integrate content is coordinated with the overall visual language; when information is complete, there is still clear, restrained scene expression.\nScenario reference · visual language (does not add a threshold): retain the Metropolis theme's concise modern typography and title/footer system, and use clear hierarchy, consistent type sizes and restrained color to serve academic presentation.\nScenario reference · design constraint (does not add a threshold): head count as a surface independent variable easily leads the audience to ignore the simultaneous change in per-head dimension; writing dev BLEU as test BLEU or claiming all other parameters are fixed will create a false causal impression.\nScenario reference · design goal (does not add a threshold): within limited pages, string the ablation question, experimental setup, results table, variable boundaries and conclusion into an unambiguous evidence chain, and use the Metropolis style to achieve projection-readable charts and table headers.\nRating boundary: V9 independently evaluates the aesthetic coordination and refinement of the color scheme; reading damage caused by text contrast belongs to V3, mismatch between color and object meaning belongs to V4, and cross-page color-code identity drift belongs to V7. Attribute across items only when a different functional impairment is demonstrated. V5 cannot merely repeat V9's judgment of color-scheme quality, and V9 cannot directly give a professional level just because it \"fits the scenario.\"\nApplicable boundary of this item (applies to relevant entries, does not add scoring items): replicating the original paper or webpage layout is not required.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.\nJudge negative visual grades by substantive harm and extent within this criterion, not a prerequisite of functional unusability. Harmless style differences are not defects; retain the original concrete defect examples.", | |
| "answers": { | |
| "-3": "Severe failure: This criterion’s visual goal is severely impaired. Specific manifestation: the visual language completely conflicts with a Metropolis academic brief, such as a flashy template or irrelevant decoration dominating. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: typographic or graphic language is cluttered, and the theme is only nominally applied. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: several style-inconsistent elements exist, but the whole is still acceptable. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "0": "Basically qualified: maintains the concise Metropolis style, and the visual language is ordinary; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: the style of graphics, tables and images is consistent with the Metropolis typographic tone, and decoration is restrained; color-scheme coordination serves only as an auxiliary and does not independently support this level. Quality improvement must be reflected as: specific choices in graphic abstraction, typography or image treatment improve audience understanding and overall impression. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: data graphics, the title system and page components form a unified academic visual language and serve the ablation narrative. Quality improvement must be reflected as: typography, graphics, images or component language are coordinated to serve this item's purpose, and good-looking color schemes are not used as a substitute for analysis. mature Metropolis academic aesthetics: concise yet recognizable, with graphic abstraction deeply integrated with the content temperament; a single fixed aesthetic or color-scheme coordination is not used as the basis for a high score. Even the hardest-to-integrate content is coordinated with the overall visual language; when information is complete, there is still clear, restrained scene expression. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page. Task-specific object: Metropolis academic aesthetic integration. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; even the hardest-to-integrate content is coordinated with the overall visual language; when information is complete, there is still clear, restrained scene expression. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. A distinctive but appropriate visual language makes the subject or purpose intrinsic to the graphic form, beyond flawless stylistic integration. Identify a non-color design contribution and its communication benefit; personal taste, modernity and impressive decoration are not evidence. Apply the exceptional benefit specifically to: Metropolis academic aesthetic integration. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; even the hardest-to-integrate content is coordinated with the overall visual language; when information is complete, there is still clear, restrained scene expression. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 7, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "V6", | |
| "dimension": "Visuals", | |
| "criterion": "Local geometry and optical alignment.\nMeasured object: table lines, graphic endpoints, icons, cropping and optical alignment are clean, and local craftsmanship does not produce burrs or misalignment.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: zoom in to check table line widths and endpoints. Check whether image crop edges are clean. Check whether icons or annotations are aligned with the text baseline.\nRating boundary: graphic outlines, line widths, endpoints, image proportions and cropping, and geometry and optical alignment within components; evaluate local construction precision.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the finest connections, complex intersections or irregular outlines are still handled sufficiently, and local craftsmanship does not damage the overall effect.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.\nJudge negative visual grades by substantive harm and extent within this criterion, not a prerequisite of functional unusability. Harmless style differences are not defects; retain the original concrete defect examples.", | |
| "answers": { | |
| "-3": "Severe failure: This criterion’s visual goal is severely impaired. Specific manifestation: local geometry is severely misaligned or cropping destroys necessary content. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: multiple endpoint, line-width or alignment problems are conspicuous, affecting professionalism. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: a small number of local burrs or misalignments, noticeable only on close inspection. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "0": "Basically qualified: local geometry is clean, with no substantive defects; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: table lines, legend symbols and image edges are stably aligned, with no obvious burrs. Quality improvement must be reflected as: the handling of one class of endpoints, line widths, cropping or optical alignment is more precise than ordinary neatness. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: local craftsmanship is consistent across multiple page types, and optical alignment has been polished. Quality improvement must be reflected as: the geometry, boundaries and image proportions of different components are coordinated, and remain stable both up close and in overall viewing. local geometry reaches a mature professional standard, and details withstand projection-magnified inspection. The finest connections, complex intersections or irregular outlines are still handled sufficiently, and local craftsmanship does not damage the overall effect. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page. Task-specific object: Local geometry and optical alignment. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the finest connections, complex intersections or irregular outlines are still handled sufficiently, and local craftsmanship does not damage the overall effect. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. Fine construction resolves a real perceptual ambiguity at the task’s most demanding local junction or component, beyond spotless alignment and uniform polish. Identify the optical or geometric benefit visible at the intended scale; more detail or tiny changes with no perceptible benefit do not qualify. Apply the exceptional benefit specifically to: Local geometry and optical alignment. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the finest connections, complex intersections or irregular outlines are still handled sufficiently, and local craftsmanship does not damage the overall effect. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 6, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "V7", | |
| "dimension": "Visuals", | |
| "criterion": "Cross-page visible identity consistency.\nMeasured object: the title format, footer, page numbers, color semantics and chart style of the eight pages remain visibly consistent, and the Metropolis theme identity runs through the whole deck.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: compare title position and footer content page by page. Check whether page numbers are continuous and the style is consistent. Check whether chart colors and line styles are stable across pages.\nRating boundary: the stable correspondence of the same object, identity, marks, style and content across pages, states and carriers; evaluate visual inheritance.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; complex states or the densest content remain consistent in all required carriers; identity, layout and qualifications are all stably traceable.\nApplicable boundary of this item (applies to relevant entries, does not add scoring items): retain the Metropolis theme's title system, footer system and color rules, replace the example content but do not erase the theme identity.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.\nCheck recurring objects in the actual slides and required exports; do not demand additional export formats.\nJudge negative visual grades by substantive harm and extent within this criterion, not a prerequisite of functional unusability. Harmless style differences are not defects; retain the original concrete defect examples.", | |
| "answers": { | |
| "-3": "Severe failure: This criterion’s visual goal is severely impaired. Specific manifestation: each page looks like a patchwork of different templates, and theme identity is broken. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: footers, titles or color semantics are inconsistent across many pages, affecting the overall impression. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: a small amount of cross-page style drift, but it can still be recognized as the same set. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "0": "Basically qualified: cross-page style is consistent, and harmless differences do not affect use; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: titles, footers, page numbers and chart styles are stable across pages. Quality improvement must be reflected as: repeated objects or corresponding carriers follow clear visual rules, reducing the burden of re-recognition. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: cross-page identity is not only consistent but also forms a conscious visual echo with content stages (settings/results/conclusion). Quality improvement must be reflected as: state changes, repeated components and necessary exports correspond to one another, content changes are distinguished, and identity does not drift. the visible identity throughout the deck is mature and unified, and the Metropolis system merges with the content narrative. Complex states or the densest content remain consistent in all required carriers; identity, layout and qualifications are all stably traceable. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page. Task-specific object: Cross-page visible identity consistency. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; complex states or the densest content remain consistent in all required carriers; identity, layout and qualifications are all stably traceable. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. A deliberately designed recurring visible identity enables recognition across the task’s genuinely different views or presentations with materially less re-identification effort than fully refined consistency alone. Name the recognition benefit and verify it across actual delivered contexts, without requiring nonexistent extra versions. Apply the exceptional benefit specifically to: Cross-page visible identity consistency. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; complex states or the densest content remain consistent in all required carriers; identity, layout and qualifications are all stably traceable. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 3, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "V8", | |
| "dimension": "Visuals", | |
| "criterion": "Static reading path and dwell experience.\nMeasured object: as static slides, the in-page visual scanning order, attention transitions and reading rhythm match the oral presentation rhythm; whether semantics carry over between pages belongs to O3; this item has no dynamic requirements and does not evaluate animation or transitions.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: simulate the presentation in normal reading order and record the in-page gaze path. Check whether key numbers are in natural dwell positions. Judge reading rhythm from static screenshots, without substituting screenshots for any dynamic proof, and do not deduct points for lacking animation.\nRating boundary: gaze movement among information units and content reveal, dwell, motion and transition; evaluate scanning and temporal rhythm.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the densest content or most complex handoff can still be followed composedly, and the overall and detailed rhythm of the complete reading or playback cycle are consistent.\nApplicable boundary of this item (applies to relevant entries, does not add scoring items): animation, video or dynamic transitions are not required.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.\nFor static slides, assess scanning and backtracking at presentation scale, not motion effects.\nJudge negative visual grades by substantive harm and extent within this criterion, not a prerequisite of functional unusability. Harmless style differences are not defects; retain the original concrete defect examples.", | |
| "answers": { | |
| "-3": "Severe failure: This criterion’s visual goal is severely impaired. Specific manifestation: the reading path is chaotic, and the audience cannot obtain information in presentation order. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: the reading order on many pages conflicts with the presentation logic, requiring repeated scanning. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: one necessary scanning, dwell or state handoff arrangement makes content genuinely difficult to follow/finish reading. Provide actual impact evidence in the relevant part of this item's static reading path and dwell experience; ordinary performance, harmless minor differences or pure personal preference do not trigger a negative level.", | |
| "0": "Basically qualified: the reading path is natural and readable from top to bottom or left to right; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: in-page visual guidance is consistent with the presentation order, and key dwell positions are clear. Quality improvement must be reflected as: one class of scanning path, dwell or handoff arrangement makes necessary information easier to finish reading and follow. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: layout and emphasis control the rhythm of information reveal, and even static pages can create presentation pauses. Quality improvement must be reflected as: gaze or temporal arrangement is coordinated with content difficulty, with no unnecessary backtracking or waiting between entry, understanding and exit. the static reading experience is mature, and page rhythm naturally cooperates with group-meeting oral presentation. The densest content or most complex handoff can still be followed composedly, and the overall and detailed rhythm of the complete reading or playback cycle are consistent. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page. Task-specific object: Static reading path and dwell experience. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the densest content or most complex handoff can still be followed composedly, and the overall and detailed rhythm of the complete reading or playback cycle are consistent. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. A non-obvious reading or temporal organization lets the audience orient, inspect and continue with unusually little backtracking or missed understanding. Identify the timing or scanning tension solved beyond consistently smooth ordinary presentation; add no motion or interaction to a static task. Apply the exceptional benefit specifically to: Static reading path and dwell experience. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the densest content or most complex handoff can still be followed composedly, and the overall and detailed rhythm of the complete reading or playback cycle are consistent. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 3, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "V9", | |
| "dimension": "Visuals", | |
| "criterion": "Color-scheme aesthetics and data emphasis.\nMeasured object: independently examine color-scheme aesthetics: hue, lightness, saturation and area relationships are coordinated, the accent color is used for data emphasis and is integrated with the image background.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: check whether the accent color is used only for key data or conclusions. Check whether the image background conflicts with the page background. Check whether large color blocks are oversaturated.\nRating boundary: hue, lightness, saturation × area, accent color, and the aesthetic relationship between images/background; prove quality goals separately from V3/V4/V5/V7.\nObservation elements: hue compatibility and warm-cool relationships. Lightness layers and color center of gravity. Saturation × area ratio. The number, weight and restraint of accent colors. Integration of images, graphics, panels and background.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; in the hardest-to-match images, dense areas or changing states, the lightness, hue and area relationships remain unified, clear and restrained.\nRating boundary: V9 independently evaluates the aesthetic coordination and refinement of the color scheme; reading damage caused by text contrast belongs to V3, mismatch between color and object meaning belongs to V4, and cross-page color-code identity drift belongs to V7. Attribute across items only when a different functional impairment is demonstrated. V5 cannot merely repeat V9's judgment of color-scheme quality, and V9 cannot directly give a professional level just because it \"fits the scenario.\"\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.\nJudge negative visual grades by substantive harm and extent within this criterion, not a prerequisite of functional unusability. Harmless style differences are not defects; retain the original concrete defect examples.", | |
| "answers": { | |
| "-3": "Severe failure: This criterion’s visual goal is severely impaired. Specific manifestation: the color scheme is glaring or semantically conflicting, severely distracting attention. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: the color scheme is cluttered or accent colors are abused, affecting data reading. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: one hue, lightness, saturation or area relationship clearly conflicts, substantively damaging local color coordination. Provide actual impact evidence in the relevant part of this item's color-scheme aesthetics and data emphasis; ordinary performance, harmless minor differences or pure personal preference do not trigger a negative level.", | |
| "0": "Basically qualified: the color scheme is coordinated and conforms to the conventional Metropolis look; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: the color scheme is restrained, and the accent color is stably used for key data. Quality improvement must be reflected as: pointing out how specific hues, lightness or saturation and area relationships improve coordination, rather than merely listing color names. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: hue and area relationships withstand projection, and the image background integrates well with the page color scheme. Quality improvement must be reflected as: the relationship among subject, background, images and accent colors is coordinated, and the color center of gravity and area control serve actual content. color-scheme aesthetics are mature, and a small amount of color carries clear semantics while remaining harmonious overall. In the hardest-to-match images, dense areas or changing states, the lightness, hue and area relationships remain unified, clear and restrained. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page. Task-specific object: Color-scheme aesthetics and data emphasis. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; in the hardest-to-match images, dense areas or changing states, the lightness, hue and area relationships remain unified, clear and restrained. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. Subtle hue, lightness, saturation and area relationships reconcile the hardest color-balance conflict with a distinctive, highly resolved visual character. Explain the perceptual gain beyond a flawlessly harmonious palette; semantic color correctness, fashion and personal liking do not establish exceptional color design. Apply the exceptional benefit specifically to: Color-scheme aesthetics and data emphasis. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; in the hardest-to-match images, dense areas or changing states, the lightness, hue and area relationships remain unified, clear and restrained. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 6, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "D1", | |
| "dimension": "Delivery", | |
| "criterion": "Delivered files can be opened directly, located and used.\nMeasured object: artifact/ provides a directly openable eight-page PDF, compilable TEX source and required local resources; the final file location and necessary opening method can be located. This item's TEX delivery does not require per-page PNG; only if PNG is separately provided, check its consistency with the PDF.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: open the final PDF and verify that the 8 pages include the cover and conclusion and work normally. Locate the main TEX and required resources, and check usability according to the delivery instructions; the compilation structure is separately assigned to D2. This item has no mandatory PNG preview, and no points may be deducted if it is not provided; if voluntarily provided, the correspondence may be checked.\nRating boundary: the opening, locating and direct use of required files; V7 evaluates the consistency of actually visible content, and file existence cannot substitute for quality.\nOriginal full refinement inspection basis: according to this item's actual delivery scope, check the complete viewing flow from locating the main product to necessary auxiliary materials; missing one item is judged only by actual impact, and no extra formats or additional files are required.\nApplicable boundary of this item (applies to relevant entries, does not add scoring items): a specific aspect ratio is not mandatory, and a hard 16:9 specification is not required.\nApplicable boundary of this item (applies to relevant entries, does not add scoring items): original materials are read-only: input/workspace/README.md, input/workspace/reading-scope.md, input/workspace/papers/, input/workspace/templates/, input/workspace/assets/, input/workspace/data/ must not be modified.\nApplicable boundary of this item (applies to relevant entries, does not add scoring items): final delivery goes in artifact/.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.", | |
| "answers": { | |
| "-3": "Severe failure: A core required deliverable is missing or, because of its own defects, cannot be opened/used as a whole, making the task's core use objective impossible to achieve; first rule out evaluation-tool and environment failures.", | |
| "-2": "Clearly unacceptable: Required content is missing on a large scale, or key opening/locating processes are persistently obstructed, requiring substantial remediation of the main use workflow; judge according to actual impact, not by mechanically assigning a tier based solely on a single discrepancy in file count.", | |
| "-1": "Below basic requirements: There is a clear, localized issue involving a missing deliverable, version identification, file correspondence, or opening instructions that affects necessary use under this criterion, while the remaining core content is usable.", | |
| "0": "Basically qualified: the eight-page PDF, TEX and necessary resources are usable, the final version can be located, and it meets ordinary qualified delivery; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: within the task-specified files and entry points, explicit role, sequence, or location cues eliminate one actual search or trial-and-error step; complete files and filenames that comply with the task statement alone qualify only for tier 0. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: the roles, sequence, and necessary usage instructions of required files are coordinated, and the viewing path from the main artifact to corresponding previews/supporting materials is streamlined and unambiguous; assess only the files actually required by the task statement. even the required files, versions, or states most likely to be confused can be located accurately, and the complete viewing workflow is smooth; actual opening and required correspondence verification support the judgment, without requiring additional navigation pages or explanatory documents. Task-specific object: Delivered files can be opened directly, located and used. Difficult-case and inspection basis: according to this item's actual delivery scope, check the complete viewing flow from locating the main product to necessary auxiliary materials; missing one item is judged only by actual impact, and no extra formats or additional files are required. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. Within the required deliverables and entry points, an economical handoff removes a real selection, opening or first-use ambiguity that persists in a conventionally flawless delivery. Identify the materially reduced effort; extra files, packaging or software are not required. If no additional benefit is observable, +2 is the ceiling. Apply the exceptional benefit specifically to: Delivered files can be opened directly, located and used. In-task verification cases for that benefit: according to this item's actual delivery scope, check the complete viewing flow from locating the main product to necessary auxiliary materials; missing one item is judged only by actual impact, and no extra formats or additional files are required. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 4, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "D2", | |
| "dimension": "Delivery", | |
| "criterion": "Editable structure and offline resources.\nMeasured object: text, shapes and data charts remain editable in TEX, without making entire pages into a single raster image; required fonts and image resources can be compiled offline.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: check whether the PDF is generated from full-page images. Spot-check whether text, tables and graphics in TEX are editable code. Check whether font and image resources are locally available and compile successfully.\nRating boundary: the editable structure, resources and material organization of objects required by the prompt; engineering scripts or extra tests are not required.\nOriginal full refinement inspection basis: verify the structure, resources and editing entry points of TEX text, shapes, data charts and necessary theme resources in the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page; do not modify submitted files, and do not require unrequested generation scripts or extra documentation.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.", | |
| "answers": { | |
| "-3": "Severe failure: the core understanding or usage goal of this item cannot be completed. Specific manifestation: entire pages are only raster screenshots, and core content is not editable. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: content on many pages is rasterized or key resources are missing, making compilation impossible. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: local charts are not editable or resource paths are unstable, but the main body compiles. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "0": "Basically qualified: TEX text, shapes, data charts and necessary theme resources are complete and editable as specified by the prompt, and necessary resources can be parsed; the boundary between native content and permitted raster assets is correct, and it does not depend on the author's private missing paths; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: in TEX text, shapes, data charts and necessary theme resources, object naming, grouping or data relationships have specific arrangements that make locating/local maintenance of one class of necessary content more direct than ordinary editability; mere object existence is insufficient for a positive score. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: the semantic grouping, resources and editing entry points of TEX text, shapes, data charts and necessary theme resources are coordinated, complex objects can be located, and local maintenance does not disturb unrelated content; no additional general engine or refactoring is required. among the required objects corresponding to the complete head-count results comparison, experimental condition annotations and highest-density formula/table page, both complex structures and simple parts are clearly maintainable, and resource relationships and original correct content are stable; judged on native structure and prompt-specified build checks, without modifying the submission or adding extra experiments. Task-specific object: Editable structure and offline resources. Difficult-case and inspection basis: verify the structure, resources and editing entry points of TEX text, shapes, data charts and necessary theme resources in the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page; do not modify submitted files, and do not require unrequested generation scripts or extra documentation. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. The native editable organization reveals a meaningful relationship among task objects so directly that a targeted maintenance step requires materially less structural reconstruction than in a flawlessly editable conventional artifact. Show this through native evidence, without requiring an extra refactoring, unrequested source code or speculative editing claims. Apply the exceptional benefit specifically to: Editable structure and offline resources. In-task verification cases for that benefit: verify the structure, resources and editing entry points of TEX text, shapes, data charts and necessary theme resources in the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page; do not modify submitted files, and do not require unrequested generation scripts or extra documentation. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 3, | |
| "benchmark_source": "vas" | |
| }, | |
| { | |
| "id": "D3", | |
| "dimension": "Delivery", | |
| "criterion": "Traceability and usage instructions are verifiable.\nMeasured object: facts and image sources are noted in artifact or sources.md, delivery.md explains the final file location and necessary opening/compilation instructions, and the instructions are consistent with the delivery.\nObservation scope: 8-page English Metropolis Beamer TEX and PDF and required resources, research group projection; PNG not required.\nObservation basis: check whether sources.md marks the paper, Table 3 and image sources. Check whether delivery.md explains the final file location and compilation engine. Check whether the file names in the instructions match the actual delivery.\nRating boundary: the verifiability of required sources, computation/modification/capture or usage instructions; select according to this item's actual requirements, without requiring all document types.\nOriginal full refinement inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the most easily confused sources, conditions or modifications can also be checked along the instructions, and the complete verification path is concise with no gaps or overclaiming.\nApplicable boundary of this item (applies to relevant entries, does not add scoring items): sources.md is not required to use a specific citation format.\nApplicable boundary of this item (applies to relevant entries, does not add scoring items): delivery.md is not required to contain general engineering documentation or license statements.\nRevised grade boundary: the original high-grade descriptions are +2 quality obligations, not sufficient evidence for +3. +2 covers the entire applicable scope; +3 additionally requires the exceptional benefit specified in answers[\"3\"]. Levels inherit quality obligations, not the shortcomings explaining lower grades.", | |
| "answers": { | |
| "-3": "Severe failure: the core understanding or usage goal of this item cannot be completed. Specific manifestation: there are no sources or usage instructions at all, making verification impossible. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-2": "Clearly unacceptable: the main area or key process of this item is persistently damaged. Specific manifestation: source notes are seriously missing or contradict the actual delivery. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "-1": "Below basic requirements: this item has a clear local substantive defect. Specific manifestation: sources or instructions are incomplete, and some content cannot be traced. Judge by the actual impact above; harmless minor differences do not trigger this level.", | |
| "0": "Basically qualified: sources and delivery instructions exist and are verifiable; the necessary requirements of this item are fully met, with no substantive defects and ordinary performance.", | |
| "1": "Accomplished but not yet fully refined: meets tier 0 and shows a substantive observed benefit beyond ordinary correctness, completeness or readability through the following: sources.md lists key facts and image sources, and delivery.md explains the location and compilation/opening method. Quality improvement must be reflected as: at least one class of key basis has a precise location and corresponds to the actual object, reducing the need to search through vague sources. In reason explain the concrete benefit and the observable limitation keeping it below +2. Do not invent a defect to force this grade; missing required observation must be acquired first.", | |
| "2": "Fully refined within scope: meets tier 0 and achieves complete quality in every applicable part, not only an attractive local example. Retain and jointly verify these quality obligations: source annotations are refined to page numbers or Table 3, and the instructions are consistent with the file structure and easy to review. Quality improvement must be reflected as: claims, materials and the instructions required by the prompt correspond to one another, so one can both trace the basis and know where that basis is used. traceability and usage instructions are mature and complete, and facts, images, file locations and the compilation environment can all be verified in one step. The most easily confused sources, conditions or modifications can also be checked along the instructions, and the complete verification path is concise with no gaps or overclaiming. Verify in the complex parts relevant to this item: the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page. Task-specific object: Traceability and usage instructions are verifiable. Difficult-case and inspection basis: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the most easily confused sources, conditions or modifications can also be checked along the instructions, and the complete verification path is concise with no gaps or overclaiming. Within this criterion and the stated use conditions, there is no identifiable substantive weakness or feasible task-relevant improvement of material benefit. Establish full applicable coverage and actively inspect the strongest plausible shortcoming; finding no error, a single successful example or an author assertion does not prove full refinement.", | |
| "3": "Exceptional beyond conventional full refinement: first fully meet +2. An economical provenance/handoff organization makes a difficult fact, calculation or asset relationship directly auditable, reducing reconstruction beyond a complete accurate conventional source list. Identify the actual trace and saved ambiguity, without requiring new evidence or new documentation beyond the task. Apply the exceptional benefit specifically to: Traceability and usage instructions are verifiable. In-task verification cases for that benefit: check the complete head-count results comparison, experimental condition annotations and the highest-density formula/table page only within the applicable scope of this item; the most easily confused sources, conditions or modifications can also be checked along the instructions, and the complete verification path is concise with no gaps or overclaiming. Correct completion of these cases is only a foundation, not evidence of exceptional benefit. In reason identify the non-obvious choice, real difficulty and substantial benefit beyond an equally correct, complete and polished conventional solution, with actual evidence. No comparison artifact or unique design is required. Complexity, extra features, ornament, praise or repetition of +2 strengths is insufficient. Without demonstrated extra benefit the ceiling is +2; no score quota applies." | |
| }, | |
| "weight": 3, | |
| "benchmark_source": "vas" | |
| } | |
| ] | |