morgan freeman
LR$^2$Bench: Evaluating Long-chain Reflective Reasoning Capabilities of Large Language Models via Constraint Satisfaction Problems
Chen, Jianghao, Wei, Zhenlin, Ren, Zhenjiang, Li, Ziyong, Zhang, Jiajun
Recent progress in o1-like models has significantly enhanced the reasoning abilities of Large Language Models (LLMs), empowering them to tackle increasingly complex tasks through reflection capabilities, such as making assumptions, backtracking, and self-refinement. However, effectively evaluating such reflection capabilities remains challenging due to the lack of appropriate benchmarks. To bridge this gap, we introduce LR$^2$Bench, a novel benchmark designed to evaluate the Long-chain Reflective Reasoning capabilities of LLMs. LR$^2$Bench comprises 850 samples across six Constraint Satisfaction Problems (CSPs) where reflective reasoning is crucial for deriving solutions that meet all given constraints. Each type of task focuses on distinct constraint patterns, such as knowledge-based, logical, and spatial constraints, providing a comprehensive evaluation of diverse problem-solving scenarios. We conduct extensive evaluation on both conventional models and o1-like models. Our experimental results reveal that even the most advanced reasoning-specific models, such as DeepSeek-R1 and OpenAI o1-preview, struggle with tasks in LR$^2$Bench, achieving an average Exact Match score of only 20.0% and 23.6%, respectively. These findings underscore the significant room for improvement in the reflective reasoning capabilities of current LLMs. The leaderboard of our benchmark is available at https://huggingface.co/spaces/UltraRonin/LR2Bench
- Asia > China > Shanghai > Shanghai (0.04)
- North America > United States > Florida > Miami-Dade County > Miami (0.04)
- Asia > Vietnam > Hanoi > Hanoi (0.04)
- (3 more...)
- Leisure & Entertainment > Sports (0.68)
- Leisure & Entertainment > Games (0.48)
- Information Technology > Artificial Intelligence > Representation & Reasoning > Constraint-Based Reasoning (1.00)
- Information Technology > Artificial Intelligence > Natural Language > Large Language Model (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks > Deep Learning (1.00)
- Information Technology > Artificial Intelligence > Cognitive Science > Problem Solving (1.00)
Morgan Freeman calls AI deepfake a 'scam' after his voice is replicated on TikTok
The Fox News contributor argued law enforcement is good for all communities to ensure public safety. With one of the most discernible voices in Hollywood, Morgan Freeman certainly has his money where his mouth is. So, it's no surprise that the revered actor took issue with a video circulating on the social platform TikTok, with a voice that was packaged as his own. "Welcome to my niece's day-in-life, narrated by me, Morgan Freeman," the video begins. The creator captioned her post in part, "Uncle Mo has been booked and busy, but i finally got him to narrate my trip!" ACTOR MORGAN FREEMAN DERIDES BLACK HISTORY MONTH: 'MY HISTORY IS AMERICAN HISTORY' Morgan Freeman rebuked a TikTok that claimed to be narrated by him.
- Media > News (0.77)
- Law Enforcement & Public Safety > Crime Prevention & Enforcement (0.57)
'World's most advanced' humanoid robot does impressions of Morgan Freeman, Elon Musk, and Donald Trump - and they're eerily realistic
It's already predicted the future, told terrible jokes, and demonstrated a range of realistic facial expressions including blinking and smiling. Now, British humanoid robot, Ameca, has been showing off its range of celebrity impressions – and they're eerily realistic. In a new video, the sophisticated machine – developed by Cornwall-based firm Engineered Arts – speaks in the style of Morgan Freeman, Elon Musk, and Donald Trump. Ameca is fitted with microphones, binocular eye mounted cameras, a chest camera and facial recognition software to interact with people. The robot has been described as the'world's most advanced' humanoid by Engineered Arts, and a'platform for human-robot interaction'. 'The aim here is to build the best expressive capabilities,' Engineered Arts says.
- North America > United States (0.64)
- Europe > Spain > Catalonia > Barcelona Province > Barcelona (0.05)
- Asia > China > Beijing > Beijing (0.05)
Revealed: The actors who would make the best Santa in a Christmas movie, according to AI - so, do you agree with its suggestions?
From Richard Attenborough in'Miracle on 34th Street' to Kurt Russell in'The Christmas Chronicles' a number of famous actors have taken on the role of Santa Claus in blockbuster hits through the years. But who would take on the leading role if Hollywood cast a new movie featuring Father Christmas? To answer this burning question, MailOnline turned to ChatGPT. While the AI bot says that casting for a dream Santa would depend on the tone and style of the film, it suggests five actors who could take on the role. So, do you agree with its star-studded suggestions?
- Media > Film (1.00)
- Leisure & Entertainment (1.00)
Ex-Google chief warns AI could displace humans for sex and love: Why would one 'need another being?'
Fox News anchor Julie Banderas reacts to the vice president's gaffe and CNN calling Dylan Mulvaney a man on'Jesse Watters Primetime.' Former chief from Google Mo Gawdat warned that artificial intelligence could lead to a "very significant redesign of love and relationships." The ex-Google X chief business officer recently appeared on an episode of the Impact Theory with Tom Bilyeu podcast, where the two discussed the future implications of AI simulating sex and relationships. "Just think about all of the illusions that we're now unable to decipher illusion from truth, right? Sex happens in the brain at the end of the day, I mean the physical side of it is not that difficult to simulate okay? But if we can convince you that this sex robot is alive or that sex experience in a virtual reality headset or an augmented reality headset is real, then there you go," he said.
- Europe > United Kingdom > England (0.07)
- Europe > Switzerland > Geneva > Geneva (0.07)
Former Google chief says AI will soon bring sex dolls to life - as he warns it will 'redesign love and relationships'
'Let's just say this is a very significant redesign of society,' said Mo Gawdat, the former chief business officer at Google's secretive R&D wing, Google X. The convergence of these technologies, as Gawdat explained on a recent podcast interview, may lead to sex dolls that seem'alive' or dating apps filled with AI'avatars.' 'If we think a few years further and think of Neuralink and other ways of connecting directly to your nervous system,' Gawdat speculated, 'why would you need another being in the first place?' Speaking on the YouTube channel for the show Impact Theory with Tom Bilyeu, Gawdat pointed out that technologists, policymakers and society at large often focus too tightly on philosophical questions that big business interests will not. 'We get lost in those conversations of'Are they alive?
- Health & Medicine > Therapeutic Area > Neurology (0.37)
- Health & Medicine > Health Care Technology (0.37)
Triple2Vec: Learning Triple Embeddings from Knowledge Graphs
Fionda, Valeria, Pirró, Giuseppe
Graph embedding techniques allow to learn high-quality feature vectors from graph structures and are useful in a variety of tasks, from node classification to clustering. Existing approaches have only focused on learning feature vectors for the nodes in a (knowledge) graph. To the best of our knowledge, none of them has tackled the problem of embedding of graph edges, that is, knowledge graph triples. The approaches that are closer to this task have focused on homogeneous graphs involving only one type of edge and obtain edge embeddings by applying some operation (e.g., average) on the embeddings of the endpoint nodes. The goal of this paper is to introduce Triple2Vec, a new technique to directly embed edges in (knowledge) graphs. Trple2Vec builds upon three main ingredients. The first is the notion of line graph. The line graph of a graph is another graph representing the adjacency between edges of the original graph. In particular, the nodes of the line graph are the edges of the original graph. We show that directly applying existing embedding techniques on the nodes of the line graph to learn edge embeddings is not enough in the context of knowledge graphs. Thus, we introduce the notion of triple line graph. The second is an edge weighting mechanism both for line graphs derived from knowledge graphs and homogeneous graphs. The third is a strategy based on graph walks on the weighted triple line graph that can preserve proximity between nodes. Embeddings are finally generated by adopting the SkipGram model, where sentences are replaced with graph walks. We evaluate our approach on different real world (knowledge) graphs and compared it with related work.
- North America > United States > Illinois > Cook County > Chicago (0.09)
- Europe > Italy > Calabria (0.04)
- Information Technology > Artificial Intelligence > Representation & Reasoning > Semantic Networks (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Statistical Learning (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Supervised Learning > Representation Of Examples (0.54)
The Best and Worst Super Bowl Ads
Two years ago, I watched every single Super Bowl, so I can say with absolute certainty that, for whatever reason, America's most-watched sports event is usually a terrible football game. This, I suspect, is one of the reasons why Americans have come to care so much about Super Bowl ads: They know that while the game will probably fall short of its hype, at least they'll see a few entertaining commercials. Last year's Super Bowl reversed that trend, with a great game surrounded by a bunch of lackluster ads. This year's Super Bowl followed suit. While the game itself was an all-time classic--easily top 10, maybe even top five--this year's ads were a poor crop. For every humorous or striking one, there were at least three others that were boastful, cloying, cringeworthy, or misguided.
Mark Zuckerberg's Virtual Assistant is Voiced by Morgan Freeman
STAFFVIRTUAL is a Business Process Outsourcing (BPO) and HR Services company with offices in California and the Philippines. Hire dedicated remote staff with us to perform your company's repetitive processes. Your team works from our modern offices in the Philippines, and is a seamless extension of your office. We're helping to create a more equitable world where nationalities matter least - and capabilities matter most.
- Asia > Philippines (0.74)
- North America > United States > California (0.39)
What were your earliest memories of going to the movies? Here are ours
Megyn Kelly says she'll ask Putin directly about allegations of election meddling Happy birthday to Morgan Freeman, who turns 80 today Jennifer Garner takes issue with new People magazine cover Chloë Grace Moretz addresses body-shaming controversy over Snow White movie Megyn Kelly says she'll ask Putin directly about allegations of election meddling What were your earliest memories of going to the movies? The Golden Age of the multiplex is in the past. Theater owners are luring a new generation with upgraded screens and snacks. And even with the rising prices, tech distractions and rude patrons, there are still many pleasures to be had at the cinema. The L.A. Times film staff reminisced about their buttered-popcorn-scented memories and how the theater-going experience (sticky floors and all) made them fall for that old cinematic magic.
- Media > Film (1.00)
- Leisure & Entertainment (1.00)
- Government (1.00)