Author(s): S Aishwarya Originally published on Towards AI. Strong data storytelling goes beyond simply visualizing numbers it uncovers the meaning behind the patterns, bringing clarity to what would otherwise be just a spreadsheet of values. Photo by Carlos Muza on Unsplash While visualization libraries like matplotlib, Plotly, and Seaborn can produce beautiful charts, they often lack one crucial feature: narrative. They leave it up to the viewer to interpret the story behind the lines and bars. That’s where Altair and the pynarrative library shine. Together, they help us not only visualize data — but actually explain it. 🔍What is Altair? Altair is a Python library for declarative data visualization that allows users to create clean, concise, and interactive charts based on the Vega-Lite grammar of graphics. You only need to provide: your data (typically a pandas DataFrame or Vega datasets) chart type (e.g., bar, line, scatter) encoding (x/y axes, color, size, etc.) optional interactivity, filtering, and tooltips Altair then renders the visualization using a JSON specification — ready for use in dashboards, notebooks, web applications, or reports. The Altair library directly integrates with pandas (for data handling) and Vega-Lite (for rendering and interactivity), making it easy for Python users to create powerful data stories without writing complex plotting code. 🔍 What is pynarrative? pynarrative is a Python library designed to automatically craft clear, insightful narrative summaries from pandas DataFrames and Altair charts. With just a few inputs: A dataset (often a time series or structured data in a DataFrame) A visualization (created using Altair) Axis labels, optional context, and your intended message pynarrative generates a well-structured textual explanation — ideal for embedding in dashboards, reports, presentations, or interactive data stories. Built to work seamlessly with pandas (for data handling) and Altair (for visual rendering), pynarrative helps bridge the gap between raw data and human-readable insights — turning visualizations into compelling narratives with minimal effort. Data Description: We’re using the cars dataset, which contains information about different car models. The main features we’ll focus on are: Horsepower: The power of the car’s engine. Miles_per_Gallon (MPG): How fuel-efficient the car is. Origin: Where the car was made (USA, Europe, or Japan). Name: The model name of the car. These features help us explore the relationship between a car’s power and fuel efficiency, and how that varies by origin. Data Cleaning & Preparation We’ll begin by automatically loading the dataset using Seaborn, then clean it for our visualizations. import pandas as pdimport seaborn as snsimport altair as altimport pynarrative as pndf = sns.load_dataset('mpg')print(df[['name', 'mpg', 'horsepower', 'origin']].head()) Output: Image by Author Cleaning Steps: Convert horsepower to numeric to handle any potential issues. Drop rows with missing values in critical fields. df['horsepower'] = pd.to_numeric(df['horsepower'], errors='coerce')df_clean = df.dropna(subset=['horsepower', 'mpg', 'origin'])print(df_clean[['name', 'mpg', 'horsepower', 'origin']].head()) Output: Image by Author Story 1: Power vs. Fuel Efficiency Let’s explore the relationship between a car’s engine power (horsepower) and its fuel efficiency (miles per gallon). By color-coding the data points based on the car’s region of origin, we gain insight into how different countries approach automotive design. chart = pn.Story(df_clean).mark_circle(size=60).encode( x='horsepower:Q', y='mpg:Q', color='origin:N', tooltip=['name', 'horsepower', 'mpg', 'origin']).add_title( "Horsepower vs MPG by Origin", "Higher horsepower often leads to lower fuel efficiency", title_color="#1a1a1a", subtitle_color="#4a4a4a").add_context( text=[ "Cars with more horsepower generally consume more fuel.", "Japanese and European models show a clear emphasis on fuel efficiency.", "This trend reveals differing consumer needs and manufacturer strategies." ], position="bottom", dx=0, color="black").render()chart Output: Image by Author This visualization reveals that American cars tend to have higher horsepower but lower fuel economy, whereas Japanese and European cars show more balance. Story 2: Regional Efficiency Trends Over Time Let’s observe how fuel efficiency (MPG) has changed over time across different regions. # Estimate year from model_year columndf_clean['year'] = df_clean['model_year'] + 1900# Compute average MPG by region and yearregional_avg = df_clean.groupby(['year', 'origin'])['mpg'].mean().reset_index()chart = pn.Story(regional_avg).mark_line(point=True).encode( x=alt.X('year:O', axis=alt.Axis(title='Year')), y=alt.Y('mpg:Q', title='Average MPG'), color='origin:N').add_title( "Average Fuel Efficiency by Region Over Time", "Trends in MPG from 1970s to 1980s", title_color="#1a1a1a", subtitle_color="#4a4a4a").add_context( text=[ "Japanese cars consistently lead in fuel efficiency.", "U.S. manufacturers ramped up efficiency post-1975.", "European models maintain a steady middle ground." ], position="bottom", dx=0, color="black").render()chart Output: Image by Author We see how regulatory changes and fuel crises influenced fuel efficiency, especially in the U.S. Story 3: Impact of the 1973 Oil Crisis Let’s annotate our chart with the 1973 Oil Crisis, a pivotal moment for car design. chart = pn.Story(regional_avg).mark_line().encode( x='year:O', y='mpg:Q', color='origin:N').add_title( "Impact of the 1973 Oil Crisis on MPG", "Shift in design philosophy after fuel shortages", title_color="#1a1a1a", subtitle_color="#4a4a4a").add_context( text=[ "The 1973 Oil Crisis increased focus on fuel efficiency worldwide.", "U.S. automakers shifted designs to improve MPG post-crisis.", "Japanese models were already MPG leaders at the time." ], position="bottom", dx=0, color="black").add_annotation( 1973, 15.5, "1973 Oil Crisis", arrow_direction='up', arrow_dx=0, arrow_dy=-1, arrow_color='red', arrow_size=50, label_color='black', label_size=14, show_point=True).render()chart Output: Image by Author This annotated visualization adds historical context, showing how global events shape industry trends. In Summary… Using pynarrative and Altair, we seamlessly transformed car performance data into engaging visual stories by: Highlighting the inverse relationship between horsepower and fuel efficiency Exploring how regional design philosophies shape fuel economy over time Annotating major historical events like the 1973 Oil Crisis to show their industry impact All of this was done using a single, intuitive interface combining pandas, Altair, and pynarrative. Once this storytelling pipeline is in place, it can be adapted to any dataset rendered through Altair from automotive to healthcare and beyond. This approach is quicker, more scalable, and more intuitive than conventional manual charting methods. Whether you’re building technical reports, dynamic dashboards, or insight-driven narratives, this serves as a reliable foundation for effective data storytelling. I would love to read your comments! Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor. Published via Towards AI

VLMs have become central to building general-purpose AI systems capable of understanding and interacting in digital and real-world settings. By integrating visual and textual data, VLMs have driven advancements in multimodal reasoning, image editing, GUI agents, robotics, and more, influencing sectors like education and healthcare. Despite this progress, VLMs still lag behind human capabilities, particularly in tasks involving 3D reasoning, object counting, creative visual interpretation, and interactive gameplay. A challenge lies in the scarcity of rich, diverse multimodal datasets, unlike the abundant textual resources available to LLMs. Additionally, multimodal data complexity poses significant training and evaluation hurdles.  Researchers at ByteDance have developed Seed1.5-VL, a compact yet powerful vision-language foundation model featuring a 532 M-parameter vision encoder and a 20 B-parameter Mixture-of-Experts LLM. Despite its efficient architecture, Seed1.5-VL achieves top results on 38 out of 60 public VLM benchmarks, excelling in tasks like GUI control, video understanding, and visual reasoning. It is trained on trillions of multimodal tokens using advanced data synthesis and post-training techniques, including human feedback. Innovations in training, such as hybrid parallelism and vision token redistribution, optimize performance. The model’s efficiency and strong reasoning capabilities suit real-world interactive applications like chatbots.  The Seed1.5-VL architecture features a vision encoder, an MLP adapter, and an LLM. Its custom vision encoder, Seed-ViT, supports native-resolution image input using 2D RoPE and processes images through 14×14 patches, followed by average pooling and an MLP. Pretraining involves masked image modeling, contrastive learning, and omni-modal alignment using images, text, and video-audio-caption pairs. The model uses a Dynamic Frame-Resolution Sampling approach for video encoding that adapts frame rates and resolutions based on content complexity, balancing efficiency and detail. This method enables effective spatial-temporal understanding within a token budget, ensuring comprehensive video representation across varied lengths and complexities.  The pre-training of Seed1.5-VL involved curating 3 trillion high-quality tokens across diverse domains. Image-text pairs from the web were filtered using CLIP scores, size/aspect ratio checks, and deduplication to reduce noise. Using domain-based sampling and duplication strategies, rare visual concepts were overrepresented to address class imbalance. Specialized datasets were added for OCR using annotated and synthetic text-rich images, charts, and tables—object grounding and counting tasks utilized bounding boxes, points, and auto-labeled web data. Additional tasks included 3D spatial understanding using depth annotations, and video understanding through multi-frame captioning, QA, and temporal grounding to support dynamic content analysis.  The evaluation highlights Seed-ViT and Seed1.5-VL’s competitive performance across vision-language tasks. Seed-ViT, despite having significantly fewer parameters, matches or outperforms larger models like InternVL-C and EVA-CLIP on zero-shot image classification tasks, showing high accuracy and robustness on datasets such as ImageNet-A and ObjectNet. Seed1.5-VL demonstrates strong capabilities in multimodal reasoning, general VQA, document understanding, and grounding. It achieves state-of-the-art benchmarks, particularly in complex reasoning, counting, and chart interpretation tasks. The model’s “thinking” mode, which incorporates longer reasoning chains, further enhances performance, indicating its strong ability in detailed visual understanding and task generalization.  In conclusion, Seed1.5-VL is a vision-language foundation model featuring a 532 M-parameter vision encoder and a 20 B-parameter Mixture-of-Experts language model. Despite its compact size, it achieves state-of-the-art results on 38 of 60 public benchmarks and excels in complex reasoning, OCR, diagram interpretation, 3D spatial understanding, and video analysis. It also performs well in agent-driven tasks like GUI control and gameplay, surpassing models like OpenAI CUA and Claude 3.7. The model shows strong generalization to tasks beyond its training scope. The study outlines its architecture, data pipeline, and training methods and identifies future directions, including enhancing tool-use and visual reasoning capabilities.  Check out the Paper and Project Page. All credit for this research goes to the researchers of this project. Also, feel free to follow us on Twitter and don’t forget to join our 90k+ ML SubReddit. The post ByteDance Introduces Seed1.5-VL: A Vision-Language Foundation Model Designed to Advance General-Purpose Multimodal Understanding and Reasoning appeared first on MarkTechPost.

With the amount of Data growing exponentially in the last few years, one of the biggest challenges has become finding the most optimal way to store various data flavors. Unlike in the (not so far) past, when relational databases were considered the only way to go, organizations now want to perform analysis over raw data – think of social media sentiment analysis, audio/video files, and so on – which usually couldn’t be stored in a traditional (relational) way, or storing them in a traditional way would require significant effort and time, which increase the overall time-for-analysis. Another challenge was to somehow stick with a traditional approach to have data stored in a structured way, but without the necessity to design complex and time-consuming ETL workloads to move this data into the enterprise data warehouse. Additionally, what if half of the data professionals in your organization are proficient with, let’s say, Python (data scientists, data engineers), and the other half (data engineers, data analysts) with SQL? Would you insist that “Pythonists” learn SQL? Or, vice-versa? Or, would you prefer a storage option that can play to the strengths of your entire data team? I have good news for you – something like this has already existed since 2013, and it’s called Apache Parquet! Parquet file format in a nutshell Before I show you the ins and outs of the Parquet file format, there are (at least) five main reasons why Parquet is considered a de facto standard for storing data nowadays: Data compression – by applying various encoding and compression algorithms, Parquet file provides reduced memory consumption Columnar storage – this is of paramount importance in analytic workloads, where fast data read operation is the key requirement. But, more on that later in the article… Language agnostic – as already mentioned previously, developers may use different programming languages to manipulate the data in the Parquet file Open-source format – meaning, you are not locked with a specific vendor Support for complex data types Row-store vs Column-store We’ve already mentioned that Parquet is a column-based storage format. However, to understand the benefits of using the Parquet file format, we first need to draw the line between the row-based and column-based ways of storing the data. In traditional, row-based storage, the data is stored as a sequence of rows. Something like this: Image by author Now, when we are talking about OLAP scenarios, some of the common questions that your users may ask are: How many balls did we sell? How many users from the USA bought a T-shirt? What is the total amount spent by customer Maria Adams? How many sales did we have on January 2nd? To be able to answer any of these questions, the engine must scan each and every row from the beginning to the very end! So, to answer the question: how many users from the USA bought T-shirt, the engine has to do something like this: Image by author Essentially, we just need the information from two columns: Product (T-Shirts) and Country (USA), but the engine will scan all five columns! This is not the most efficient solution – I think we can agree on that… Column store Let’s now examine how the column store works. As you may assume, the approach is 180 degrees different: Image by author In this case, each column is a separate entity – meaning, each column is physically separated from other columns! Going back to our previous business question: the engine can now scan only those columns that are needed by the query (Product and country), while skipping scanning the unnecessary columns. And, in most cases, this should improve the performance of the analytical queries. Ok, that’s nice, but the column store existed before Parquet and it still exists outside of Parquet as well. So, what is so special about the Parquet format? Parquet is a columnar format that stores the data in row groups Wait, what?! Wasn’t it complicated enough even before this? Don’t worry, it’s much easier than it sounds Let’s go back to our previous example and depict how Parquet will store this same chunk of data: Image by author Let’s stop for a moment and explain the illustration above, as this is exactly the structure of the Parquet file (some additional things were intentionally omitted, but we will come soon to explain that as well). Columns are still stored as separate units, but Parquet introduces additional structures, called Row group. Why is this additional structure super important? You’ll need to wait for an answer for a bit :). In OLAP scenarios, we are mainly concerned with two concepts: projection and predicate(s). Projection refers to a SELECT statement in SQL language – which columns are needed by the query. Back to our previous example, we need only the Product and Country columns, so the engine can skip scanning the remaining ones. Predicate(s) refer to the WHERE clause in SQL language – which rows satisfy criteria defined in the query. In our case, we are interested in T-Shirts only, so the engine can completely skip scanning Row group 2, where all the values in the Product column equal socks! Image by author Let’s quickly stop here, as I want you to realize the difference between various types of storage in terms of the work that needs to be performed by the engine: Row store – the engine needs to scan all 5 columns and all 6 rows Column store – the engine needs to scan 2 columns and all 6 rows Column store with row groups – the engine needs to scan 2 columns and 4 rows Obviously, this is an oversimplified example, with only 6 rows and 5 columns, where you will definitely not see any difference in performance between these three storage options. However, in real life, when you’re dealing with much larger amounts of data, the difference becomes more evident. Now, the fair question would be: how does Parquet “know” which row group to skip/scan? Parquet file contains metadata This means that every Parquet file contains “data about data” – information such as minimum and maximum values in a specific column within a certain row group. Furthermore, every Parquet file contains a footer, which keeps the information about the format version, schema information, column metadata, and so on. You can find more details about Parquet metadata types here. Important: In order to optimize the performance and eliminate unnecessary data structures (row groups and columns), the engine first needs to “get familiar” with the data, so it first reads the metadata. It’s not a slow operation, but it still requires a certain amount of time. Therefore, if you’re querying the data from multiple small Parquet files, query performance can degrade, because the engine will have to read metadata from each file. So, you should be better off merging multiple smaller files into one bigger file (but still not too big :)… I hear you, I hear you: Nikola, what is “small” and what is “big”? Unfortunately, there is no single “golden” number here, but for example, Microsoft Azure Synapse Analytics recommends that the individual Parquet file should be at least a few hundred MBs in size. What else is in there? Here is a simplified, high-level illustration of the Parquet file format: Image by author Can it be better than this? Yes, with data compression Ok, we’ve explained how skipping the scan of the unnecessary data structures (row groups and columns) may benefit your queries and increase the overall performance. But, it’s not only about that – remember when I told you at the very beginning that one of the main advantages of the Parquet format is the reduced memory footprint of the file? This is achieved by applying various compression algorithms. I’ve already written about various data compression types in Power BI (and the Tabular model in general) here, so maybe it’s a good idea to start by reading this article. There are two main encoding types that enable Parquet to compress the data and achieve astonishing savings in space: Dictionary encoding – Parquet creates a dictionary of the distinct values in the column, and afterward replaces “real” values with index values from the dictionary. Going back to our example, this process looks something like this: Image by author You might think: why this overhead, when product names are quite short, right? Ok, but now imagine that you store the detailed description of the product, such as: “Long arm T-Shirt with application on the neck”. And, now imagine that you have this product sold million times…Yeah, instead of having million times repeating value “Long arm…bla bla”, the Parquet will store only the Index value (integer instead of text). Run-Length-Encoding with Bit-Packing – when your data contains many repeating values, Run-Length-Encoding (RLE) algorithm may bring additional memory savings. Can it be better than THIS?! Yes, with the Delta Lake file format Ok, what the heck is now a Delta Lake format?! This is the article about Parquet, right? So, to put it in plain English: Delta Lake is nothing else but the Parquet format “on steroids”. When I say “steroids”, the main one is the versioning of Parquet files. It also stores a transaction log to enable tracking all changes applied to the Parquet file. This is also known as ACID-compliant transactions. Since it supports not only ACID transactions, but also supports time travel (rollbacks, audit trails, etc.) and DML (Data Manipulation Language) statements, such as INSERT, UPDATE and DELETE, you won’t be wrong if you think of the Delta Lake as a “data warehouse on the data lake” (who said: Lakehouse). Examining the pros and cons of the Lakehouse concept is out of the scope of this article, but if you’re curious to go deeper into this, I suggest you read this article from Databricks. Conclusion We evolve! Same as we, the data is also evolving. So, new flavors of data required new ways of storing it. The Parquet file format is one of the most efficient storage options in the current data landscape, since it provides multiple benefits – both in terms of memory consumption, by leveraging various compression algorithms, and fast query processing by enabling the engine to skip scanning unnecessary data. Thanks for reading! The post Parquet File Format – Everything You Need to Know! appeared first on Towards Data Science.

Nature, Published online: 14 May 2025; doi:10.1038/s41586-025-09001-2Neurons in the rodent dorsomedial prefrontal cortex encode a flexible internal model of emotion by linking directly experienced and inferred associations with aversive experiences.

May 13, 2025Ravie LakshmananCyber Espionage / Malware The North Korea-linked threat actor known as Konni APT has been attributed to a phishing campaign targeting government entities in Ukraine, indicating the threat actor's targeting beyond Russia. Enterprise security firm Proofpoint said the end goal of the campaign is to collect intelligence on the "trajectory of the Russian invasion." "The group's interest in Ukraine follows historical targeting of government entities in Russia for strategic intelligence gathering purposes," security researchers Greg Lesnewich, Saher Naumaan, and Mark Kelly said in a report shared with The Hacker News. Konni APT, also known as Opal Sleet, Osmium, TA406, and Vedalia, is a cyber espionage group that has a history of targeting entities in South Korea, the United States, and Russia. It's operational since at least 2014. Attack chains mounted by the threat actor often involve the use of phishing emails to distribute malware called Konni RAT (aka UpDog) and redirect recipients to credential harvesting pages. Proofpoint, in an analysis of the threat group published in November 2021, assessed TA406 to be one of several actors that make up the activity publicly tracked as Kimsuky, Thallium, and Konni Group. The latest set of attacks documented by the cybersecurity company entails the use of phishing emails that impersonate a fictitious senior fellow at a think tank called the Royal Institute of Strategic Studies, which is also a non-existent organization. The email messages contain a link to a password-protected RAR archive that's hosted on the MEGA cloud service. Opening the RAR archive using a password mentioned in the message body launches an infection sequence that's engineered to conduct extensive reconnaissance of the compromised machines. Specifically, present within the RAR archive is a CHM file that displays decoy content related to former Ukrainian military leader Valeriy Zaluzhnyi. Should the victim click anywhere on the page, a PowerShell command embedded within the HTML is executed to reach out to an external server and download a next-stage PowerShell payload. The newly launched PowerShell script is capable of executing various commands to gather information about the system, encode it using Base64-encoding, and send it to the same server. "The actor sent multiple phishing emails on consecutive days when the target did not click the link, asking the target if they had received the prior emails and if they would download the files," the researchers said. Proofpoint said it also observed an HTML file being directly distributed as an attachment to the phishing messages. In this variation of the attack, the victim is instructed to click on an embedded link in the HTML file, resulting in the download of a ZIP archive that includes a benign PDF and a Windows shortcut (LNK) file. When the LNK is run, it executes Base64-encoded PowerShell to drop a Javascript Encoded file called "Themes.jse" using a Visual Basic Script. The JSE malware, in turn, contacts an attacker-controlled URL and runs the response from the server via PowerShell. The exact nature of the payload is currently not known. Furthermore, TA406 has been spotted attempting to harvest credentials by sending fake Microsoft security alert messages to Ukrainian government entities from ProtonMail accounts, warning them of suspicious sign-in activity from IP addresses located in the United States and urging them to verify the login by visiting a link. While the credential harvesting page has not been recovered, the same compromised domain is said to have been used in the past to collect Naver login information. "These credential harvesting campaigns took place prior to the attempted malware deployments and targeted some of the same users later targeted with the HTML delivery campaign," Proofpoint said. "TA406 is very likely gathering intelligence to help North Korean leadership determine the current risk to its forces already in the theatre, as well as the likelihood that Russia will request more troops or armaments." "Unlike Russian groups who have likely been tasked with gathering tactical battlefield information and targeting of Ukrainian forces in situ, TA406 has typically focused on more strategic, political intelligence collection efforts." The disclosure comes as the Konni group has been linked to a sophisticated multi-stage malware campaign targeting entities in South Korea with ZIP archives containing LNK files, which run PowerShell scripts to extract a CAB archive and ultimately deliver batch script malware capable of collecting sensitive data and exfiltrating it to a remote server. The findings also dovetail with spear-phishing campaigns orchestrated by Kimsuky to target government agencies in South Korea by delivering a stealer malware capable of establishing command-and-control (C2 or C&C) communications and exfiltrating files, web browser data, and cryptocurrency wallet information. According to South Korean cybersecurity company AhnLab, Kimsuky has also been observed propagating PEBBLEDASH as part of a multi-stage infection sequence initiated via spear-phishing. The trojan was attributed by the U.S. government to the Lazarus Group in May 2020. "While the Kimsuky group uses various types of malware, in the case of PEBBLEDASH, they execute malware based on an LNK file by spear-phishing in the initial access stage to launch their attacks," it said. "They then utilize a PowerShell script to create a task scheduler and register it for automatic execution. Through communication with a Dropbox and TCP socket-based C&C server, the group installs multiple malware and tools including PEBBLEDASH." Konni and Kimsuky are far from the only North Korean threat actors to focus on Seoul. As recently as March 2025, South Korean entities have been found to be at the receiving end of another campaign carried out by APT37, which is also referred to as ScarCruft. Dubbed Operation ToyBox Story, the spear-phishing attacks singled out several activists focused on North Korea, per the Genians Security Center (GSC). The first observed spear phishing attack occurred on March 8, 2025. "The email contained a Dropbox link leading to a compressed archive that included a malicious shortcut (LNK) file," the South Korean company said. "When extracted and executed, the LNK file activated additional malware containing the keyword 'toy.'" The LNK files are configured to launch a decoy HWP file and run PowerShell commands, leading to the execution of files named toy03.bat, toy02.bat, and toy01.bat (in that order), the last of which contains shellcode to launch RoKRAT, a staple malware associated with APT37. RokRAT is equipped to collect system information, capture screenshots, and use three different cloud services, including pCloud, Yandex, and Dropbox for C2. "The threat actors exploited legitimate cloud services as C2 infrastructure and continued to modify shortcut (LNK) files while focusing on fileless attack techniques to evade detection by antivirus software installed on target endpoints," Genians said. Found this article interesting? Follow us on Twitter  and LinkedIn to read more exclusive content we post. SHARE    