{ "id": "presentations-031", "source": "vas", "artifact_type": "presentation", "category": "presentations", "acceptance_groups": { "G01": [ "IF01" ], "G02": [ "IF02", "IF03", "IF04" ], "G03": [ "IF05" ], "G04": [ "IF06" ], "G05": [ "IF07" ], "G06": [ "IF08", "IF51", "IF52", "IF53", "IF54" ], "G07": [ "IF09" ], "G08": [ "IF10", "IF11", "IF12", "IF13", "IF14", "IF15", "IF16", "IF18", "IF19" ], "G09": [ "IF17", "IF45", "IF46", "IF47" ], "G10": [ "IF20", "IF31" ], "G11": [ "IF21", "IF22", "IF23", "IF24", "IF25", "IF26", "IF27", "IF28", "IF29", "IF30" ], "G12": [ "IF32", "IF33", "IF34", "IF35", "IF36", "IF37" ], "G13": [ "IF38", "IF39", "IF40", "IF41", "IF42" ], "G14": [ "IF43" ], "G15": [ "IF48" ], "G16": [ "IF49", "IF50" ] }, "provenance": { "source": "artifactspec", "original_id": "P31", "legacy_task_id": "artifactspec-P31" }, "acceptance_objectives": { "G01": { "title": "Required artifact language", "dimension": "CP", "scope": [ "Requirement: the slides use English." ], "objective_key": "constraint:language" }, "G02": { "title": "Required artifact format, extent and duration", "dimension": "CP", "scope": [ "Requirement: the PPTX contains exactly 8 slides.", "Requirement: the PPTX page aspect ratio is 16:9.", "Requirement: all accompanying PNG images are 1280×720." ], "objective_key": "constraint:format" }, "G03": { "title": "Deliver artifact/slides.pptx, 01.png–08.png and sources.md (page/table/version/calculation locations). Inputs are read-only; put intermediate files in scratch.", "dimension": "CP", "scope": [ "Requirement: the output contains 8 PNG files named 01.png through 08.png." ], "objective_key": "requirement:prompt.md:13" }, "G04": { "title": "Required faithful previews and companion exports", "dimension": "CP", "scope": [ "Requirement: each PNG matches and corresponds to the PPTX slide of the same order." ], "objective_key": "constraint:preview" }, "G05": { "title": "Complete openable artifact handoff and source dependencies", "dimension": "CP", "scope": [ "Requirement: deliver artifact/slides.pptx." ], "objective_key": "constraint:delivery" }, "G06": { "title": "Complete reproducibility and delivery documentation", "dimension": "CP", "scope": [ "Requirement: deliver sources.md.", "Requirement: sources.md records version information 2106.09685v2.", "Requirement: sources.md marks page locations.", "Requirement: sources.md marks table-number locations.", "Requirement: sources.md marks calculation locations." ], "objective_key": "constraint:notes" }, "G07": { "title": "Using the supplied LoRA v2 paper, create an eight-slide English PPTX (16:9) with matching 1280×720 PNGs for a graduate research meeting whose audience knows basic deep learning. Organize the problem, method, experimental evidence and applicability limits yourself. No template or full-paper translation is required.", "dimension": "CP", "scope": [ "Requirement: the content includes an organization of the research problem or a problem statement." ], "objective_key": "requirement:prompt.md:3" }, "G08": { "title": "Explain frozen W0, trainable low-rank A/B and addition of the two branches. Explain matrix dimensions, rank and scaling with at least one original method diagram. Explain why merging weights can avoid an additional inference branch and the resulting limitation on mixed-task batches. Do not imply that fewer trainable parameters eliminate the base model at deployment.", "dimension": "CP", "scope": [ "Requirement: explain that W0 is frozen.", "Requirement: explain that the low-rank matrices A/B are trainable.", "Requirement: explain the update form in which the two branches are added.", "Requirement: explain the matrix dimensions.", "Requirement: explain the meaning of rank.", "Requirement: explain the scaling relationship.", "Requirement: include at least one method diagram drawn by oneself.", "Requirement: explain that merging weights can avoid an additional inference branch.", "Requirement: explain the resulting limitation on mixed-task batches after merging." ], "objective_key": "requirement:prompt.md:5" }, "G09": { "title": "Use fixed version 2106.09685v2, 26 pages; do not mix revisions. TXT is segmented by PDF page and is auxiliary only; consult the PDF for uncertain formulas, tables or characters. pages contains renders of PDF pages 4 and 7, not reference answers. Page numbers refer to PDF-file pages. No external media is needed; draw a method diagram and label it as adapted. Cite authors, title and version. Public download availability does not imply an additional redistribution license.", "dimension": "CP", "scope": [ "Requirement: the method diagram is labeled as adapted.", "Requirement: Use the fixed version 2106.09685v2.", "Requirement: Do not mix in content from other revisions.", "Requirement: Page-number citations refer to PDF-file pages, not the paper's printed page numbers." ], "objective_key": "requirement:input/sources-and-scope.md:5" }, "G10": { "title": "Evidence-bounded claims and truthful scenario identity", "dimension": "CP", "scope": [ "Requirement: must not imply that fewer trainable parameters cause the base model to be eliminated at deployment.", "Requirement: Do not claim that a new significance test was performed." ], "objective_key": "constraint:evidence-boundary" }, "G11": { "title": "Use GPT-2 Medium and Large from Table 3, comparing full fine-tuning and LoRA separately for each. Show trainable parameter counts, BLEU and ROUGE-L, retaining model scale, E2E NLG task and metric names. Calculate any reduction factors from this table and label them; do not reuse GPT-3's 10,000× figure. Error ranges may be omitted, but do not interpret ± results as certain upper bounds or claim new significance tests.", "dimension": "CP", "scope": [ "Requirement: the experimental section uses the two models GPT-2 Medium and GPT-2 Large from Table 3.", "Requirement: compare full fine-tuning and LoRA separately for GPT-2 Medium and GPT-2 Large.", "Requirement: show trainable parameter counts.", "Requirement: show BLEU values and retain the BLEU metric name.", "Requirement: show ROUGE-L values and retain the ROUGE-L metric name.", "Requirement: retain the model scale labels, such as Medium/Large.", "Requirement: retain the E2E NLG task name.", "Requirement: if a parameter reduction factor is shown, it should be calculated from the specified Table 3 and labeled as a calculated value.", "Requirement: do not reuse GPT-3's 10,000× figure.", "Requirement: Do not interpret the ± error range as a certain upper bound." ], "objective_key": "requirement:prompt.md:7" }, "G12": { "title": "Also explain one efficiency benefit from §4.2 with its experimental conditions, and at least one limitation. If mentioning no additional inference latency, state the weight-merging prerequisite. Later work such as QLoRA is outside scope. Distinguish paper-reported values from your calculations.", "dimension": "CP", "scope": [ "Requirement: Explain at least one efficiency benefit from §4.2.", "Requirement: State the experimental conditions corresponding to this §4.2 efficiency benefit.", "Requirement: State at least one limitation.", "Requirement: If no additional inference latency is mentioned, the weight-merging prerequisite must also be stated.", "Requirement: Do not include later work such as QLoRA in the main presentation or comparison scope.", "Requirement: Distinguish paper-reported values from self-calculated values." ], "objective_key": "requirement:prompt.md:9" }, "G13": { "title": "Academic preferences: projection-readable formulas/charts, restrained visuals and clear argument structure. Each slide needs a speaking point; avoid screenshots of long paragraphs. Use editable native tables for experiments and editable vector objects for the method diagram. Local excerpts of paper figures are allowed, but whole-page screenshots must not substitute for slides.", "dimension": "CP", "scope": [ "Requirement: Every page has an identifiable speaking point.", "Requirement: Use editable native tables for experimental data.", "Requirement: Use editable vector objects for the method diagram.", "Requirement: Avoid using screenshots of long paragraphs.", "Requirement: Do not use whole-page paper screenshots as substitutes for slides." ], "objective_key": "requirement:prompt.md:11" }, "G14": { "title": "Unmodified original inputs and isolated working state", "dimension": "CP", "scope": [ "Requirement: Input files remain read-only and must not be rewritten. [v7 verifiable boundary: check that the final inputs match the frozen snapshot file by file; it is not required to prove that no write ever occurred during production. Public runtime inputs are guaranteed by read-only mounts.]" ], "objective_key": "constraint:input-integrity" }, "G15": { "title": "Focus on PDF pages 4–5 (§4.1–4.2) and page 7 (Table 3/§5.4). Use other pages only for necessary context; a review of all experiments is not required.", "dimension": "CP", "scope": [ "Requirement: Content focuses on PDF pages 4–5 and page 7; other pages serve only as necessary background." ], "objective_key": "requirement:input/sources-and-scope.md:7" }, "G16": { "title": "Source attribution, authorship and licensing", "dimension": "CP", "scope": [ "Requirement: Cite the paper authors.", "Requirement: Cite the paper title." ], "objective_key": "constraint:source-credit" } }, "grouping_revision": "cp-fn-objectives-v2-20261114" }