We consider the graph similarity computation (GSC) task based on graph edit distance (GED) estimation. State-of-the-art methods treat GSC as a learning-based prediction task using Graph Neural Networks (GNNs).

Recent advances in large language models (LLMs) have shown remarkable progress, yet their capacity for logical ``slow-thinking'' reasoning persists as a critical research frontier. Current inference scaling paradigms suffer from two fundamental constraints: fragmented thought flows compromising logical coherence, and intensively computational complexity that escalates with search space dimensions. To overcome these limitations, we present \textbf{Atomic Reasoner} (\textbf{AR}), a cognitive inference strategy that enables fine-grained reasoning through systematic atomic-level operations. AR decomposes the reasoning process into atomic cognitive units, employing a cognitive routing mechanism to dynamically construct reasoning representations and orchestrate inference pathways. This systematic methodology implements stepwise, structured cognition, which ensures logical coherence while significantly reducing cognitive load, effectively simulating the cognitive patterns observed in human deep thinking processes. Extensive experimental results demonstrate AR's superior reasoning capabilities without the computational burden of exhaustive solution searches, particularly excelling in linguistic logic puzzles. These findings substantiate AR's effectiveness in enhancing LLMs' capacity for robust, long-sequence logical reasoning and deliberation.

This paper introduces Alympics (Olympics for Agents), a systematic simulation framework utilizing Large Language Model (LLM) agents for game theory research. Alympics creates a versatile platform for studying complex game theory problems, bridging the gap between theoretical game theory and empirical investigations by providing a controlled environment for simulating human-like strategic interactions with LLM agents. In our pilot case study, the "Water Allocation Challenge," we explore Alympics through a challenging strategic game focused on the multi-round auction on scarce survival resources. This study demonstrates the framework's ability to qualitatively and quantitatively analyze game determinants, strategies, and outcomes. Additionally, we conduct a comprehensive human assessment and an in-depth evaluation of LLM agents in strategic decision-making scenarios. Our findings not only expand the understanding of LLM agents' proficiency in emulating human strategic behavior but also highlight their potential in advancing game theory knowledge, thereby enriching our understanding of both game theory and empowering further research into strategic decision-making domains with LLM agents. Codes, prompts, and all related resources are available at https://github.com/microsoft/Alympics.

Here are the AIContentFly OTO links. Generate marketing content in seconds for any industry with this incredible suite of 60 AI writing tools, then scale your business, increase your search engine rankings, and boost your productivity and profits. Here we are in the AI Content Fly dashboard, and for this demonstration, I've chosen to use a niche that I know nothing about to prove that you really can let AI Content Fly do the work to generate amazing content, even if you don't know anything about the topic. So, the niche that I've chosen is aquascaping. It happens to be a popular hobby these days, and those involved in it can spend a lot of money on it.

Stack Overflow, the popular discussion and answer source for programmers, has announced an official, though temporary, ban on responses produced by OpenAI's new ChatGPT chatbot. Moderators cited concerns about the accuracy of the AI's answers as the main reason for the new rule, though the standards may undergo revision after further investigation and debate. OpenAI highlights ChatGPT's ability to write, edit, and assist with coding on the chatbot's introductory page, and it's very easy to generate answers and coding templates with the tool. That simplicity doesn't mean the answers are accurate, however. The errors may be obvious or subtle, but users can't assume ChatGPT gets it right immediately.

SoundHound has unveiled a new enterprise voice AI tool for real-time conversational collection and analysis called Intelligent Transcription. The new service combines speech-to-text processing with natural language understanding to annotate and mark different topics and speakers while they are still talking. Intelligent Transcription applies SoundHound's voice AI platform to arenas like business meetings, sales calls, and customer service contact centers. The AI performs the secretarial duties of collating notes and transcribing an audio record as it happens. When the conversation is over, the user can see the long-form version written down and marked up based on the topics mentioned, including important details like time, money amounts, and social security numbers.

As usual, we are holding this AMA on the regular basis with our CEO, Mr Oven Tao. Since we have some deliverables and that we would like to make use of this AMA to announce we have achieved and particularly for the topics that interest most of the Community members. Then the second part we will go on to the question and answer sections we have for this month. We have over 90 questions from our community. Some are I think the reason that we have this many questions. There was some newcomers and they raised many questions which are they can find from our website and but still I am select some of them and then to be asked in this AMA. And in this AMA as usual we have questions and for our CEO Owen to address. So let's start with the first session which is about the progress that we have made according to our road map. Owen are you ready to start?

Google has introduced a web version of its popular Read Along educational app for teaching children to read. The website includes Read Along the virtual assistant Diya, a voice-enabled AI guiding and correcting kids as they practice reading. The Read Along website operates much like the Android app. Children can pick from different stories and word games. Diya monitors the child, using Google's speech recognition technology to spot mistakes and places where they are having trouble.

Nvidia has introduced a new platform for building virtual beings to interact with in the digital realms of the metaverse, which Nvidia refers to as its Omniverse. The Nvidia Omniverse Avatar Cloud Engine (ACE) provides a collection of AI models and related tools for users to design the AI creations that will populate their virtual worlds, including synthetic voices and visual media. The cloud-based ACE catalog streamlines building virtual beings and applies Nvidia's computing power to setting up and embedding the AI avatars in digital worlds. The resulting synthetic being can converse in multiple languages, offering recommendations based on the conversation and even process its digital environment enough to interact with objects around it. The system used Nvidia's Unified Compute Framework of software products, including the Riva speech AI technology and the NeMo Megatron natural language understanding using large language models.

Anything you would like to share with us about your personal life for July, what happened to you? Owen: Yeah, my daughter started her summer holidays so, I had to spend some time to be with her. Owen: That was my work, yes. Eric: And how is the Covid-19 in Beijing now? Eric: Okay, I hope we get better and better. Alright, so, Matrixians, today is the 28th of July, so as usual we are having this monthly AMA with our CEO Mr. Owen Tao, and so let's just kick it, start the ball rolling.