AllYouCanTech

**Ferret-UI Lite: A Milestone in On-Device GUI Engagement**

With only 3 billion parameters, Ferret-UI Lite either matches or exceeds the benchmark performance of models that are up to 24 times larger. Here are the specifics.

### A Brief Background on Ferret

In December 2023, a group of 9 researchers released a paper titled “FERRET: Refer and Ground Anything Anywhere at Any Granularity.” Within it, they introduced a multimodal large language model (MLLM) capable of comprehending natural language references to specific components of an image. Since then, Apple has published subsequent papers that broaden the Ferret model family, including Ferretv2, Ferret-UI, and Ferret-UI 2.

Ferret-UI variants particularly built upon the original FERRET functionalities, remedying the limitations of general-domain MLLMs. The foundational Ferret-UI paper underscored that despite significant progress in MLLMs, they frequently have difficulty effectively understanding and interacting with user interface (UI) displays. Ferret-UI was created to better interpret mobile UI screens, including “any resolution” to enhance details and utilize advanced visual features.

### Ferret-UI Lite

Recently, Apple enriched the Ferret-UI lineup with a paper titled “Ferret-UI Lite: Insights from Developing Compact On-Device GUI Agents.” Ferret-UI was founded on a 13B-parameter model aimed at mobile UI comprehension and fixed-resolution screenshots. In contrast, Ferret-UI Lite is a streamlined model crafted to operate on-device while competing with much larger GUI agents.

The researchers observed that traditional GUI agent methodologies often depend on extensive foundation models because of their robust reasoning and planning abilities. However, these models are usually too substantial and resource-intensive for effective on-device functionality. Consequently, they created Ferret-UI Lite, a variant with 3 billion parameters, comprising crucial elements driven by insights on training compact language models.

Ferret-UI Lite harnesses:

– Authentic and synthetic training datasets from various GUI sectors.
– On-the-fly cropping and zooming techniques to enhance understanding of specific GUI segments.
– Supervised fine-tuning and reinforcement learning strategies.

The outcome is a model that either closely resembles or even surpasses rival GUI agent models that possess up to 24 times its parameter count.

### Innovative Techniques

The infrastructure of Ferret-UI Lite encompasses real-time cropping and zooming methods. The model makes an initial prediction, crops around that prediction, and subsequently re-predicts within the cropped area, compensating for its restricted capacity to process extensive numbers of image tokens.

Another significant advancement is the capability to generate its own training data. The researchers established a multi-agent system that directly interacts with live GUI platforms to create synthetic training examples on a large scale. This system captures the nuances of real-world interaction, including mistakes and unforeseen states, which are hard to replicate with pristine, human-annotated data.

Notably, while Ferret-UI and Ferret-UI 2 evaluated themselves with iPhone screenshots and other Apple interfaces, Ferret-UI Lite was trained and assessed in Android, web, and desktop GUI settings, employing benchmarks such as AndroidWorld and OSWorld. This decision likely mirrors the accessibility of reproducible, large-scale GUI-agent test scenarios.

### Performance and Applications

The researchers discovered that although Ferret-UI Lite excelled in short-horizon, low-complexity tasks, it did not perform as robustly in more intricate, multi-step interactions, a trade-off anticipated due to the limitations of a compact, on-device model. However, Ferret-UI Lite provides a local and private agent capable of autonomously interacting with app interfaces based on user demand, thereby enhancing user experience without relying on cloud processing.

To explore further about the study, including detailed benchmark analyses and outcomes, follow this link.

### ‘Dreamer’ in iOS 26 is a Stunning Reinvention of the Default iPhone Ringtone

In the continually changing world of mobile technology, ringtones have largely become secondary, eclipsed by the vast array of features that smartphones present today. Nevertheless, Apple has made a significant push to revitalize this element of the iPhone experience with the launch of new ringtones in iOS 26.

With this latest update, users of the iPhone have access to seven all-new ringtones, most of which are inventive variations of the cherished default ‘Reflection’ ringtone. Among these, the highlight is indisputably ‘Dreamer,’ which has quickly garnered favor among numerous users.

#### The Nostalgic Appeal of Ringtones

Thinking back to the pre-smartphone days, ringtones were a vital aspect of mobile phones, often acting as a means of personal expression. As smartphones advanced, the relevance of ringtones declined, yet Apple’s choice to unveil new options in iOS 26 underscores an acknowledgment of their cultural importance.

#### A New Perspective on a Classic

The ‘Reflection’ ringtone has reached iconic status since it first appeared, and Apple’s decision to reimagine it instead of merely presenting entirely different sounds is praiseworthy. Each of the six new variations captures the spirit of the original while incorporating distinctive elements that enrich the auditory experience.

‘Dreamer’ shines with its charming lo-fi spin, offering a revitalizing interpretation of the recognizable tones. The additional options, including ‘Surge’ and ‘Reflected,’ also provide enjoyable listening experiences, demonstrating Apple’s dedication to excellence in their audio design.

#### Instructions to Change Your Ringtone

For those excited to try out the new ringtones, altering your iPhone’s ringtone is straightforward. Head to Settings, then select Sounds & Haptics, and tap on Ringtone to explore the new selections.

#### Conclusion

With the rollout of new ringtones in iOS 26, Apple has effectively reignited interest in this often-neglected feature. Whether you favor the nostalgic ‘Reflection’ or the innovative sounds of ‘Dreamer,’ there’s something available for everyone to appreciate. Have you tested any of iOS 26’s new ringtones yet? Which one do you like best?

**Safeguarding Your Family’s Privacy in the Digital Era**

In the current digital environment, securing your family’s privacy is more important than ever. The rise of data brokers—businesses that gather and sell personal details—has made privacy protection a pressing issue for numerous households. With hundreds of such entities in operation, manually removing your family’s information from the web can be overwhelming and labor-intensive.

**The Difficulty of Data Elimination**

Even with the legal framework that permits data brokers to trade personal information, individuals frequently encounter obstacles in opting out of these databases. Each data broker follows its own protocols, often intentionally complicated and tedious, making it hard for individuals to effectively request the deletion of their data. Even after managing to opt out, there is no assurance that the data won’t reappear, as brokers can obtain new information at any time.

**Streamlining the Process with Incogni**

Luckily, there are options to ease this procedure. One such option is Incogni, a data removal tool created by the team behind Surfshark. Incogni manages the heavy lifting by contacting over 250 data brokers and people search platforms on your behalf, insisting on the elimination of personal information for you and your family members.

The service not only aids in the initial removal of data but also consistently tracks for any re-emergence of your information in new databases, issuing takedown requests as necessary. This forward-thinking strategy significantly mitigates the risk of spam, junk mail, and possible identity theft for your family.

**Exclusive Offer for Family Safety**

Incogni presents a family plan that extends protection to you and up to four additional family members. Each user can add multiple contact details, ensuring thorough coverage. Currently, a promotional offer is available allowing users to receive a 55% discount on the annual plan by applying the code “9to5Mac.” Moreover, Incogni provides a 30-day money-back guarantee, enabling you to test the service without risk.

In summary, taking proactive measures to protect your family’s privacy is vital in an era where personal data is a coveted asset. Utilizing services like Incogni can streamline the data removal process and ongoing oversight, offering peace of mind in a progressively intricate digital landscape.

The activity-tracking platform Strava is introducing a fresh set of activity categories, encompassing several that have been highly requested by its user community. Here’s what’s updated.

Strava is incorporating five new activity types into its platform, which include:

– Basketball
– Cricket
– Dance
– Padel
– Volleyball

This enhancement is designed to offer its “180M+ global users more precise methods to document the activities they already enjoy.” Until now, users engaged in any of these five activities had to categorize them under the all-encompassing “Workout” option, but they will now be able to monitor their performance with greater accuracy, in addition to the more than 50 sports types that are already available.

Strava emphasizes that padel is a particularly significant inclusion due to its swift worldwide expansion, boasting around 30 million amateur players. The company also mentioned that it intends to introduce more sports and activities in the future, though specific information regarding which or the timeline for their release has not been disclosed.

Are you a Strava user? Did you notice any of the new categories? Share your thoughts in the comments.