The development of large language models (LLMs) has significantly advanced artificial intelligence (AI) across various fields. Among these advancements, mobile GUI agentsdesigned to perform tasks autonomously on smartphonesshow considerable potential. However, evaluating these agents poses notable challenges. Current datasets and benchmarks often rely on static frame evaluations, which provide snapshots of app interfaces for agents to predict the next action. This method falls short of simulating the dynamic and interactive nature of real-world mobile tasks, creating a gap between tested capabilities and actual performance. Additionally, existing platforms tend to restrict app diversity, task complexity, and real-time interaction, underscoring the need for a more comprehensive evaluation framework.In response to these challenges, researchers from CUHK, vivo AI Lab, and Shanghai Jiao Tong University have introduced the Android Agent Arena (A3), a platform designed to improve the evaluation of mobile GUI agents. A3 provides a dynamic evaluation environment with tasks that mirror real-world scenarios. The platform integrates 21 commonly used third-party apps and includes 201 tasks ranging from retrieving online information to completing multi-step operations. Additionally, A3 incorporates an automated evaluation system leveraging business-level LLMs, which reduces the need for manual intervention and coding expertise. This approach aims to close the gap between research-driven development and practical applications for mobile agents.Key Features and Advantages of A3A3 is built on the Appium framework, facilitating seamless interaction between GUI agents and Android devices. It supports a broad action space, ensuring compatibility with agents trained on diverse datasets. Tasks are categorized into three typesoperation tasks, single-frame queries, and multi-frame queriesand are divided into three levels of difficulty. This variety enables a thorough assessment of an agents capabilities, from basic navigation to complex decision-making.The platforms evaluation mechanism includes task-specific functions and a business-level LLM evaluation process. Task-specific functions use predefined criteria to measure performance, while the LLM evaluation process employs models like GPT-4o and Gemini for autonomous assessment. This combination ensures accurate evaluations and scalability for a growing number of tasks.Insights from Initial TestingThe researchers tested various agents on A3, including fine-tuned models and business-level LLMs, yielding the following insights:Challenges in Dynamic Evaluations: While agents performed well in static evaluations, they faced difficulties in A3s dynamic environment. For instance, tasks requiring multi-frame queries often resulted in low success rates, highlighting the challenges of real-world scenarios.Role of LLMs in Evaluation: The LLM-based evaluation achieved 8084% accuracy, with cross-validation reducing errors significantly. However, complex tasks occasionally required human oversight to ensure accuracy.Common Errors: Observed errors included incorrect click coordinates, redundant actions, and difficulties in self-correction. These issues underscore the need for agents capable of learning adaptively and understanding context.ConclusionAndroid Agent Arena (A3) offers a valuable framework for evaluating mobile GUI agents. By providing a diverse set of tasks, an extensive action space, and automated evaluation systems, A3 addresses many limitations of existing benchmarks. The platform represents a step forward in aligning research advancements with practical applications, enabling the development of more capable and reliable AI agents. As AI continues to evolve, A3 sets a strong foundation for future innovations in mobile agent evaluation.Check out the Paper and Project Page. All credit for this research goes to the researchers of this project. Also,dont forget to follow us onTwitter and join ourTelegram Channel andLinkedIn Group. Dont Forget to join our60k+ ML SubReddit. FREE UPCOMING AI WEBINAR (JAN 15, 2025): Boost LLM Accuracy with Synthetic Data and Evaluation IntelligenceJoin this webinar to gain actionable insights into boosting LLM model performance and accuracy while safeguarding data privacy.The post Meet Android Agent Arena (A3): A Comprehensive and Autonomous Online Evaluation System for GUI Agents appeared first on MarkTechPost.

Evaluating the real-world applicability of large language models (LLMs) is essential to guide their integration into practical use cases. One key challenge in assessing LLMs is their tendency to exploit fixed datasets during testing, leading to inflated performance metrics. Static evaluation frameworks often fail to determine a models ability to adapt to feedback or provide clarifications, resulting in evaluations that do not reflect real-world scenarios. This gap in assessment methods necessitates a dynamic and iterative framework to test models under evolving conditions.Traditionally, evaluation methods like LLM-as-a-Judge rely on fixed datasets and static benchmarks to measure performance. While these approaches often correlate better with human judgments than lexical matching techniques, they suffer from biases, including verbosity preference and inconsistent scoring across iterations. They also fail to evaluate models in multi-turn interactions, where adaptability and iterative improvement are critical. As a result, these traditional methods struggle to capture a holistic understanding of an LLMs capabilities.Researchers from KAIST, Stanford University, Carnegie Mellon University, and Contextual AI have introduced LLM-AS-AN-INTERVIEWER, a novel framework for evaluating LLMs. This approach mimics human interview processes by dynamically modifying datasets to generate tailored questions and providing feedback on model responses. The interviewer LLM adapts its questions based on the evaluated models performance, fostering a detailed and nuanced assessment of its capabilities. Unlike static methods, this framework captures behaviors such as response refinement and the ability to address additional inquiries effectively.The framework operates in three stages: problem setup, feedback and revision, and follow-up questioning. Initially, the interviewer creates diverse and challenging questions by modifying benchmark datasets. During the interaction, it provides detailed feedback on the models responses and poses follow-up questions that test additional aspects of its reasoning or knowledge. This iterative process culminates in generating an Interview Report, which compiles performance metrics, error analysis, and a comprehensive summary of the models strengths and limitations. The report offers actionable insights into the models real-world applicability and adaptability.Experiments using the MATH and DepthQA datasets demonstrate the frameworks efficacy. For MATH, which focuses on arithmetic reasoning, models like GPT-4o achieved % initial problem-solving accuracy of 72%. This accuracy increased to 84% through iterative feedback and interaction, highlighting the frameworks ability to enhance model performance. Similarly, DepthQA evaluations, which emphasize open-ended queries, revealed the effectiveness of follow-up questions in uncovering models knowledge gaps and improving their responses. For instance, the adaptability metric for GPT-3.5 showcased a marked improvement of 25% after iterative interactions, reflecting the models ability to refine answers based on feedback.The framework also addresses critical biases prevalent in LLM evaluations. Verbosity bias, a tendency to favor longer responses, diminishes as interactions progress, with a significant drop in correlation between response length and scores. Further, self-enhancement bias, where models favor their responses during evaluation, is minimized through dynamic interactions and comparative scoring. These adjustments ensure consistent and reliable evaluation outcomes across multiple runs, with standard deviations decreasing as feedback iterations increase.LLM-AS-AN-INTERVIEWER offers a robust solution to data contamination, a major LLM training and evaluation concern. The framework mitigates contamination risks by dynamically modifying benchmark questions and introducing novel follow-ups. For example, models trained on contaminated datasets showed significantly higher performance in static settings but aligned more closely with uncontaminated models when evaluated dynamically. This result underscores the frameworks ability to distinguish between genuine model capabilities and artifacts of training data overlap.In conclusion, LLM-AS-AN-INTERVIEWER represents a paradigm shift in evaluating large language models. Simulating human-like interactions and dynamically adapting to model responses provide a more accurate and nuanced understanding of their capabilities. The frameworks iterative nature highlights areas for improvement and enables models to demonstrate their adaptability and real-world applicability. With its robust design and comprehensive analysis, this framework has the potential to set a new standard for LLM evaluation, ensuring that future models are assessed with greater precision and relevance.Check out the Paper. All credit for this research goes to the researchers of this project. Also,dont forget to follow us onTwitter and join ourTelegram Channel andLinkedIn Group. Dont Forget to join our60k+ ML SubReddit. FREE UPCOMING AI WEBINAR (JAN 15, 2025): Boost LLM Accuracy with Synthetic Data and Evaluation IntelligenceJoin this webinar to gain actionable insights into boosting LLM model performance and accuracy while safeguarding data privacy.The post This AI Paper Introduces LLM-as-an-Interviewer: A Dynamic AI Framework for Comprehensive and Adaptive LLM Evaluation appeared first on MarkTechPost.

LatestMachine LearningMonth in 4 Papers (December 2024) 0 like January 3, 2025Share this postAuthor(s): Ala Falaki, PhD Originally published on Towards AI. This member-only story is on us. Upgrade to access all of Medium.This series of posts is designed to bring you the newest findings and developments in the NLP field. Ill delve into four significant research papers each month, offering a comprehensive summary. Be sure to visit my blog regularly or subscribe to my newsletter for monthly updates. Lets dive in!Also, heres to an amazing 2025! (Just a bit early!) Large Language Monkeys: Scaling Inference Compute with Repeated Sampling [paper]This paper proposes scaling the inference time as another dimension for scaling NLP models. They suggest that using the LLM to generate only one sample as the response is ineffective. The success of repeated sampling depends on two key factors: coverage and precision. Coverage refers to how many problems we can solve as we take more samples, while precision focuses on whether we can accurately choose the correct answer from the set of generated samples.One of the open challenges is to have an automatic evaluation to measure the success rate. A method like a unit test is useful for finding a correct sample from a pool of candidates, but there are tasks like math problems that we dont have a proper way Read the full blog for free on Medium.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 asponsor. Published via Towards AITowards AI - Medium Share this post

Happy new year! If you've resolved to become a better runner in 2025, now is the time to get going.But that might not be easy. With the weather downright frigid in many parts of the country, it's tempting to just curl up with some hot cocoa and enjoy some Netflix. Spring races are just around the corner, though.Whether you're looking to run your first 5K or earn another Abbott World Major Marathon star, training can't stop for the cold and snow. Getting through the next few months will take dedication, discipline and the right gear.I'm speaking from experience. I spent last winter training for the Boston Marathon. Now I'm digging out my favorite thermals as I get ready to run the TCS London Marathon on April 27.Winter running gear can sometimes get pretty pricey, but scrimping on your key layers could leave you cold and miserable. The key is to look for deals and get the best bang you can for your buck.Here's a look at some of my favorite winter and other top running gear. It's a fact: Runners love socks. FeeturesRunners love socks Upgrade your inbox Get cnet insider From talking fridges to iPhones, our experts are here to help make the world a little less complicated. When I was a kid, I used to groan when I'd find socks in my Christmas stocking. Now I get excited.Runners don't cut corners when it comes to socks. They can make the difference between finishing a race happy and limping to the end with bloody feet.That's why there's always more than one pair of Feetures exchanged at my running team's annual gift swap. Priced at about $18 a pair, Feetures' colorful and long-lasting socks are my go-to for training and racing. Specifically, the "Elite" versions of the socks offer targeted compression that helps protect and support your feet as you run, without feeling too heavy.When temperatures drop below freezing, I reach for my pair of Smartwool Cold-Weather Run crew socks. When it comes to keeping your feet warm and dry without adding a lot of bulk, you just can't beat wool.At $24, they're worth the money. Like Smartwool's other products I've used, they've managed to survive several wash cycles without shrinking as wool socks can be prone to do. Just a few pairs of the running shoes I've tried recently. Bree Fowler/CNETFor great shoes, leave it to the professionalsI could tell you what shoes I ran the TCS New York City Marathon in (Saucony's Endorphin Pro 4) or what my favorite daily trainers are (New Balance Fresh Foam X 1080v14), but it doesn't really matter.The truth is, the best shoe for you is going to depend on you, your feet and your needs. A size 9 in one brand may fit completely differently than in another. Shoes that your best friend, or even a world-class marathoner, thinks are amazing may feel torturous for you to walk in.If you're shopping for yourself, head to your local running store and get some expert advice. The fitting technology used by many stores has come a long way. In 2023, I got an up-close look at the system used by Fleet Feet, a national chain of running stores. It uses artificial intelligence to compare a 3D scan of your feet to those of millions of other customers, along with the store's inventory. Combine that tech with suggestions from a store employee and you're bound to walk away with a pair of shoes you love. Kanes are fun and comfortable. KaneFor this reason, shoes make a very bad gift, but I've got a better idea.I was happy with my marathon sneakers, but I couldn't get them off fast enough after I crossed the finish line. Once I did, I slipped on a pair ofKane recovery shoes. Yes, they look a lot like Crocs, and at $80 a pair they're a lot more expensive, but I think they're well worth the money.My swollen feet had the room they needed to breathe, while the little nodes of the relatively firm footbed massaged them as I walked the eight blocks to the subway. The shoes slipped on and off easily, and thankfully they didn't rub raw marks on my feet through the journey home.The shoes are available in a variety of fun colors and patterns, including licensed versions for alumni and fans of two dozen colleges and universities. Each pair also comes with a handy carabiner that you can use to clip them to the outside of a bag.Light brightWhether you run in the morning or the evening, it's dark outside this time of year. Lights and reflective gear are a must, both to make sure cars see you and that you see all those cracks in the sidewalk and all those unfortunately placed piles of dog poop.I've got a few sets of lights, but my favorite remains the Tracer2 light-up vest from Noxgear. Its rechargeable LED lights flash in a rainbow of fun colors. They can be seen a quarter mile away and last up to 20 hours on a single charge.At $51, it's not cheap, and I'd recommend adding on the optional Tracer Lamp ($32) that clips on the front. It's super bright and can be aimed down to help you spot cracks in the sidewalk and avoid blinding fellow runners who might try to look directly at you.The vest, which includes reflective straps, fits me perfectly and doesn't slide around like a less expensive version I used previously that was made by another company.My teammates would be upset if I didn't mention Noxgear's 39G wearable Bluetooth speaker ($56). It kicks out a surprising amount of sound for its size. Instead of using headphones or earbuds that could block out the sound of an approaching car or another threat, you can use the speaker to listen to your tunes while also staying aware of what's going on around you.Reflective clothing can also help drivers see you. If you're in the market for a winter running coat, think about getting one with reflective elements. Feetures, mentioned above, also has new reflective crew socks. They offer the same cushioning and performance as the company's other models but are designed to show up in a car's headlights.Fashionable and functionalI grew up in Michigan, so it takes a lot for me to get cold. I ran the New York City Marathon in my team singlet and a trusty pair of compression shorts from Flipbelt, even though it was 38 degrees outside when I started the race. That said, I finally pulled out my pair of Flipbelt thermal leggings ($65) this past week.Flipbelt made its name by producing stretchy running belts designed to hold everything you might need for a long run. I've had problems in the past with a belt stretching out and riding down in the middle of a race.The shorts and leggings solve that problem by building that same storage into compression bottoms that stay in place. They offer enough storage to hold your phone, keys, water flask and all the fuel and snacks you need to get through a long race.I found the leggings a little tough to pull on, but once you get them in place they don't move and they provide catsuitlike support, even if you load down the pockets. The thermal version kept me warm without adding a lot of weight. I'm looking forward to wearing these for my next race.Whether it's to shade my face from the sun or keep my ears warm, I almost always wear a hat when I run. Usually, it's a baseball cap, but when the mercury drops I'll reach for one of my fleecier options.Sprints, known for its wacky designs, makes my favorite ultralight and supercomfortable baseball-style hats. They've got some cool, new winter options that I'm looking forward to trying out.In the meantime, one of my new favorite hats comes from Turtle Fur. This merino wool cap ($58) is super light and stays on when I run, and it keeps my cold-prone ears toasty.Speaking of merino wool, Smartwool knows a few things about dressing for winter. It collaborated with shoemaker Altra to create a collection of winter running clothes complete with a matching pair of smartwool-lined Altra shoes.The women's collection comes in a beautiful eggplant, black and charcoal color combination and includes a long-sleeve top, a vest and a half-zip pullover, among other items. It launched earlier this year as anREI exclusive, but now you can find it on Smartwool's website.The shoes weren't available to test, but the clothes kept me toasty on a day when temperatures dropped into the mid 20s. The layers gave me the range of motion I needed to run, without adding a lot of bulk. I'm sure I'll be pulling these out a lot this winter. The Adidas Dunamis. AdidasShades for every budgetJust because it's cold doesn't mean it isn't bright outside. Good shades are still a must. Goodr's line of active sunglasses ($25) remain the best budget option. They block harmful UVA and UVB rays, don't slip down your nose when it gets sweaty, and come in a slew of fun colors and frames. They don't last forever, particularly if you buy the ones with mirrored lenses, which scratch easily.I recently tried a couple of pairs from Tifosi. At about $35, they cost a little bit more, but so far they seem a bit more durable. They're not heavy, but they feel sturdy. My teenage daughter instantly fell in love with my pair of the Sanctum models, and I've yet to get them back, but she hasn't broken them either. I've been wearing the Swank model.For people looking to go upscale and look very cool doing it, the Adidas Dunamis ($210) feature crystal-clear, reflective, wraparound lenses that are light as a feather and will make you feel like a professional athlete. Just a heads-up that they're meant for larger faces. If you're a small person like me, you may find them a little big.Other great gearRunners need to stay hydrated even when it's cold outside, and every runner would be happy to receive a great water bottle.My favorite right now is Nathan's ExoShot 2.0 handheld 14-ounce bottle. It's a soft flask that's held in a fabric case. There's an adjustable strap that goes around your hand, so you don't really have to cling to it while you run. It's easy to squeeze and drink from.Best of all, it doesn't leak, and you can throw it in the dishwasher. I'd use this bottle a lot more if my daughter hadn't also stolen it for her own runs.Gels, chews and other kinds of running fuel are also worthwhile. I'm a fan of Skratch Labs energy chews. They taste less like chemicals and more like gummy bears. Some of my favorite products come from UnTapped, a maple syrup company based in you guessed it Vermont. Its coffee-spiked maple syrup athletic fuel gives me the zip I need without including chemicals I can't pronounce.When I reach for a traditional gel, it's generally a Maurten. While other brands offer a Baskin-Robbins-like variety of flavors, Maurten comes in just one, which doesn't really taste like anything but isn't super sweet. It takes some getting used to, but I've found it to be easy on the tummy.

We'll show you some easy steps to keep your Wi-Fi network shielded from unwanted eyes.

Apple CEO Tim Cook has joined a chorus of Big Tech executives making sizeable donations to President-elect Donald Trumps inauguration. The news comes via a scoop from Axios as we approach Trumps inauguration on January 20. Cook will personally make a $1 million donation to Trumps inaugural committee, the report says. Last month, Cook visited Mar-a-Lago in Florida for a meeting with Trump. Tesla CEO Elon Musk was also spotted at the dinner table with Trump and Cook. Cook, a proud Alabama native, believes the inauguration is a great American tradition, and is donating to the inauguration in the spirit of unity, the sources said. The company is not expected to give.Meta CEO Mark ZuckerbergvisitedMar-a-Lago to meet with Trump at the end of November. Google CEO Sundar Pichai met withTrump in December. Trump has also met with Jeff Bezos and many other high-profile executives in the Big Tech industry and beyond. The likes of Meta, Amazon, and Uberhave donated $1 million to Trumps inauguration fund. OpenAIs Sam Altmanalso madea $1 million donation. Outside of Silicon Valley, companies including Ford, Toyota, GM, Goldman Sachs, Bank of America, Coinbase, and more have donated millions to the Trump inauguration fund. Cookmet withandearned praise from Trump multiple timesduringhis first term, andother tech leadersare reportedly trying to imitate that strategy during Trumps second term. In October,Cook spokewith Trump about Apples ongoing battles in the European Union. Trump promised Cook he would not let the EU take advantage of American companies if he were to get elected.Meta this week shook up its policy team ahead of Trump taking office later this month. Nick Clegg stepped down as Metas global head of policy after six years, with Joel Kaplan taking over. Kaplan worked under President George W. Bush and has close ties to Trump and incoming VP JD Vance. Follow Chance:Threads,Bluesky,Instagram, andMastodon.Add 9to5Mac to your Google News feed. FTC: We use income earning auto affiliate links. More.Youre reading 9to5Mac experts who break news about Apple and its surrounding ecosystem, day after day. Be sure to check out our homepage for all the latest news, and follow 9to5Mac on Twitter, Facebook, and LinkedIn to stay in the loop. Dont know where to start? Check out our exclusive stories, reviews, how-tos, and subscribe to our YouTube channel

Image: SEGA2024 is now over and that means the "Year of Shadow" and all of the celebrations related to the "Ultimate Lifeform" are winding down.Sega originally announced this annual campaign in April 2024 and followed it up with all sorts of announcements, merchandise, and toys. Sonic fans got new Shadow-themed LEGO, toys, comics, and promotional events.And then in October, Sonic Generations returned with a whole new campaign starring Shadow - there was even some DLC thrown in featuring the voice work of Keanu Reeves.The year ended on a high note with the launch of Sonic the Hedgehog 3 in cinemas - placing the spotlight on Shadow once more. The movie has now banked more than $210 million in its opening weeks and a fourth outing is reportedly on the way. Luigi (out of) time!The best 3D Sonic game..? YesChaos Control!Oh Shadow, what have they done to you?Get to know the "Ultimate Lifeform"All of this has got us wondering what our community thought about the year of Shadow. Did you enjoy it as much as we did? What did you think of it compared to Nintendo's year of Luigi? What game character would you like to see celebrated next?Vote in our poll and leave your thoughts in the comments.What score would you give the "Year of Shadow"? (32 votes)10 - It was the ultimate year!22%9 - Shadow's celebrations were excellent25%8 - I really enjoyed it16%7 - I enjoyed most of the celebrations13%6 - It wasn't bad6%5 - It was average3%4 - It could have been so much more...3%3 - Shadow's not exactly my favourite 0%2 - Yeah, it wasn't for me3%1 - I wasn't interested in any of it9% 'Technically' starting in February, but you get the pictureShare:01 Liam is a news writer and reviewer for Nintendo Life and Pure Xbox. He's been writing about games for more than 15 years and is a lifelong fan of Mario and Master Chief. Hold on there, you need to login to post a comment...Related ArticlesFeature: 15 Best-Looking Switch Games Of 2024Hello, my prettiesSega President Shuji Utsumi On Sonic The Hedgehogs Next Major OutingIt must be a "great game"Sonic The Hedgehog 'Metal Sonic' First 4 Figures Statue Revealed, Here's Another LookIf you know, you knowFeature: The Best Hidden Gems And Underrated Switch Games Of 2024Some shinies that may have slipped through your fingersJim Carrey "Open To The Idea" Of Playing Robotnik In Future Sonic MoviesIvo or Gerald, though?

It's going to be an action-packed year.To celebrate the new year, Nintendo's UK branch has shared a list of games coming to the Switch in 2025.Read the full article on nintendolife.com

LGs latest pre-CES 2025 product announcement continues the big-screen theme thats already shaping up to be one of the key trends of the show. Only this time rather than the big pictures coming from king-sized TVs, theyre coming from eye-catchingly small projectors. In fact, in the case of LGs new CineBeam S, theyre coming from what LG claims is the smallest ultra short throw 4K projector on the market.Despite weighing in at just 2.5kg and measuring just 110(w) x 160(d) x 160(h)mm, the CineBeam S (AKA the PU615U) is capable of producing pictures as large as 100-inch across from a throw distance of just a few inches. Whats more, its RGB laser-powered projection system is claimed to deliver its 3840x2160-resolution images with an impressive 500 ANSI Lumens of brightness and even more promising 450,000:1 contrast ratio.Featuring a minimalistic metallic finish, the CineBeam S blends into modern interiors while its rich ... [+] and immersive sound is claimed to come courtesy of its built-in stereo sound system with Dolby Atmos.Photo: LG Electronics USAThe CineBeam S is able to apply these impressive specifications for such a small projector to HDR signals (both the HDR10 and HLG formats are supported) as well as SDR, and is claimed to be capable of reproducing 154 percent of the DCI-P3 color space used in the mastering of most HDR content.As well as being exceptionally small, the CineBeam S reinforces its lifestyle appeal by featuring a minimalistic metallic finish, and features Auto Screen Adjustment, Wall Color Adjustment and Screen Scaling/Shifting features to make it as easy as possible to set up in any room you want to use it in.Its got built-in stereo speakers and Dolby Atmos decoding too, so that you dont have to always find an external audio system to use with it on movie nights. In fact, you can watch films and TV shows on it without even needing to connect an external source thanks to its built-in LG WebOS smart system, with its support for all the worlds most popular video streaming services.MORE FOR YOUThough if you DO want to use an external source such as a 4K Blu-ray player or games console, the projector carries an HDMI port as well as supporting wireless playback from both Android and iOS devices.The PF600U, with its 3-in-1 functionality, is a CES Innovation Award honoree, Photo: LG Electronics USALGs other new CES projector announcement is the PF600U, which despite its rather bland name is actually a pretty unique proposition. LG describes it as a versatile 3-in-1 lifestyle companion for everyday spaces a description it lives up to by combining its core projector functionality with the ability to also function as an LED mood lamp and a Bluetooth music speaker.Its design reflects its unusual flexibility, too, with its stand-pole style which has clearly been inspired by floor lamps making it easy to integrate into virtually any space.Its stereo speaker setup manages to incorporate passive radiators despite the projectors compact form, helping it deliver a richer, more rounded and so more immersive sound, while the LED mood lamp feature is capable of enhancing your rooms ambiance with nine colours and five brightness levels.The PF600U's 110-degree tilting head and auto screen adjustment function makes it easy to set up and ... [+] watch pretty much anywhere.Photo: LG Electronics USAThe PF600Us increased focus on versatility means it isnt as well specified for pictures as the CineBeam S. Its LED optical system only supports a full HD (1920x1080) resolution rather than 4K, its brightness is rated at 300 ANSI lumens rather than 500, and its contrast ratio is quoted at 150,000:1 rather than 450,000:1. It does still support HDR10 and HLG playback, though, and since its not a UST model, it is rated as being capable of delivering a 120-inch picture 20 inches more than the CineBeam S. Picture set up is made easy, too, by an auto screen adjustment function and the ability of the projector to tilt through 110 degrees on its pole mount.As with the CineBeam S, the PF600U also carries LGs webOS smart system, meaning that again you can stream movies and TV shows into the projector rather than having to attach an external source. Though also like the CineBeam S, the PF600Us 7.5kg, 290(w) c 290(d) x 983(h)mm form does carry an HDMI port, as well as supporting wireless content sharing from Android or iOS devices.LG hasnt announced pricing for its two new home entertainment projectors at the time of writing, but if this emerges during the CES next week Ill update this story accordingly.Related ReadingLG Unveils Three New OLED Gaming Monitors Including Worlds First 5K2K Bendable ModelLG Actually Puts Its Jaw-Dropping Transparent 77-Inch OLED TV On SaleLG Gets Serious About LCD TVs With New Premium QNED Evo Range

There is fresh evidence the once-mighty innovator is losing market share in more areas