{"paper_id":"2512.14735","title":"PyFi: Toward Pyramid-like Financial Image Understanding for VLMs via Adversarial Agents","authors":["Yuqun Zhang","Yuxuan Zhao","Sijia Chen"],"paper_date":"2025-12-11","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2512.14735v1","page_url":"https://thequant.space/flowcharts/pyfi-toward-pyramid-like-financial-image-understanding-for/","scores":{"empirical_rigor":7,"math_complexity":3.5,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper proposes PyFi, a framework that includes the synthetic PyFi-600K dataset and a multi‑agent adversarial mechanism (PyFi-adv) to enable VLMs to perform pyramid‑like financial image understanding via progressive question chains.","quote":"This paper proposes PyFi, a novel framework for pyramid-like financial image understanding that enables vision language models (VLMs) to reason through question chains in a progressive, simple-to-complex manner.","summary":"This paper proposes PyFi, a novel framework for pyramid-like financial image understanding that enables vision language models (VLMs) to reason through question chains in a progressive, simple-to-complex manner. At the core of PyFi is PyFi-600K, a dataset comprising 600K financial question-answer…","strategy":null},"assumptions":{"status":"none_found","items":[]},"datasets":{"status":"available","items":[{"name":"PyFi-600K","provider":null,"kind":"synthetic","access":"public","period":null,"universe":null,"quote":"We release PyFi-600K, a large-scale finance dataset comprising 600 thousand (K) samples, each consisting of a question-answer pair, constructed within a pyramid structure that encompasses six levels of finance-specific visual cognition."}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"unclear"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"forecast-accuracy","findings":[{"finding":"Fine‑tuning Qwen2.5‑VL‑3B and Qwen2.5‑VL‑7B on the pyramid‑structured question chains improves average accuracy by 19.52% and 8.06%, respectively, on the PyFi‑600K dataset.","kind":"performance","quote":"PyFi‑CoT related models improves the 3B model by 19.52% points and the 7B model by 8.06% points over their local baselines."},{"finding":"Across 15 pre‑trained VLMs, average accuracy drops from 71.80% at level 1 (Perception) to 32.95% at level 6 (Decision Support), showing difficulty with higher‑level financial reasoning.","kind":"performance","quote":"the average accuracy across all models decreases from 71.80% at level 1 (Perception) to 32.95% at level 6 (Decision Support)."},{"finding":"GLM‑4.5V achieves the highest overall accuracy of 74.75% but its accuracy falls to 46.15% at level 6, highlighting challenges in financial decision‑making.","kind":"performance","quote":"GLM-4.5V achieves the highest overall accuracy of 74.75%, demonstrating its strong performance across various capability levels. However, even this model’s accuracy drops to 46.15% at level 6."},{"finding":"Error analysis reveals that strengthening the Calculation Analysis (CA) capability level is key to improving VLM performance in financial decision‑making.","kind":"statistical","quote":"reveals that the key to improving VLMs’ performance in the financial domain lies in strengthening their capabilities at the CA (Calculation Analysis) level."},{"finding":"The pyramid‑like structure and question‑chain dataset enable back‑tracking of errors from high‑level decisions to lower‑level sub‑questions, improving interpretability.","kind":"other","quote":"PyFi‑600K enables us to trace the problem‑solving chain from DS to LR and eventually to PP, thereby diagnosing how an error in the DE leads to the final mistake."}],"reported_performance":null,"out_of_sample":{"method":"unclear","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":25,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":true,"code_url":"https://github.com/AgenticFinLab/PyFi","note":null},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":1,"extracted_at":"2026-11-14T17:19:20Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}