AI For Kids Course

AI For Kids Course — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

  • Xara Designer Pro+

    Xara Designer Pro+

    Xara Designer Pro+ is an image editing program incorporating photo editing and vector illustration tools created by British software company Xara. Xara Xtreme LX was an early open source version for Linux. The Windows version was previously sold under the names Xara Studio, Xara X and Xara Xtreme, and traces its origin in the late 1980s to a title called ArtWorks for the Acorn Archimedes line of computers using RISC OS. There is a pro version called Xara Designer Pro (formerly Xara Xtreme Pro). The current commercial version of Xara Photo & Graphic Designer runs only on Windows, although Xara documents can be edited in a web browser on any platform using the Xara Cloud service. Versions up to 4.x can be run on Linux using Wine. == History == ArtWorks, the predecessor of Xara Photo and Graphic Designer, was developed on Acorn Archimedes and Risc PC 32-bit RISC computers running RISC OS by Computer Concepts during the late 1980s. The first version, developed for Microsoft Windows was initially called Xara Studio. It was licensed to Corel Corporation before wide-scale public availability, and from 1995 to 2000 was released as CorelXARA. Corel ceded the licensing rights back to Xara in 2000. The first Xara X version released in 2000 by its original owner. The next version, Xara X¹, was released in 2004. Xara Xtreme was released in 2005. In November 2006, Xara Xtreme PRO (an enhanced version of Xara Xtreme) was released. Xara Xtreme 3.2 and Xtreme Pro 3.2 were released in May 2007. 3.2 Pro included Xara3D, and both versions had more robust typography. In April 2008, Xara Xtreme 4.0 was released. Xara Xtreme and Xara Xtreme Pro 5.1 were released in June 2009. Features included more text-area enhancements, content-aware scaling of bitmap images, improved file import and export, master-page (repeated) objects, an object gallery (replacing the layer gallery), website-creation tools, and multi-stage graduated transparency. In June 2010, Xara Photo & Graphic Designer 6 and Xara Designer Pro 6 were released. Xtreme was renamed Photo & Graphic Designer, and Xtreme Pro was renamed Designer Pro. In May 2011, Xara Photo & Graphic Designer 7 and Xara Designer Pro 7 were released. Features included "magic" photo erase, user interface improvements to docking galleries and snapping alignment, and (in Pro) new webpage and website-design features. In May 2012, Xara Photo & Graphic Designer 2013 and Xara Designer Pro X (v8) were released. Xara Photo & Graphic Designer 9 was released in May 2013. In July of that year, Xara Designer Pro X9 was released. Xara Photo & Graphic Designer 10 was released on 16 July 2014, and Xara Designer Pro X10 on 23 July. Xara Photo & Graphic Designer 11 was released on 29 June 2015, and Xara Designer Pro X11 was released the following month. In 2016, the delivery model was changed to an update service which can be renewed annually. Users are entitled to any updates released while the update service is active. The first update-service updates were in May 2016 for Xara Photo & Graphic Designer, and July 2016 for Xara Designer Pro X. == Features == Xara Photo & Graphic Designer is known for its usability and fast renderer. It provides a fully anti-aliased display, advanced gradient fill, and transparency tools. Among vector editors, Xara Photo & Graphic Designer is considered to be fairly easy to learn, with similarities to CorelDRAW and Inkscape in terms of interface. Alongside the vector illustration tools, Xara Photo & Graphic Designer also includes an integrated photo tool offering manual and automatic photo enhance, cropping, adjustment of brightness levels, red-eye fix, 'magic' erase, photo healing, color and background erase, panoramas and content aware resizing. Designer Pro includes a wider range of tools for other design tasks including the creation of web pages and websites, and text and page layout tools for DTP with the aim of providing a single solution for all graphic and web design tasks.

    Read more →
  • Text-to-image personalization

    Text-to-image personalization

    Text-to-Image personalization is a task in deep learning for computer graphics that augments pre-trained text-to-image generative models. In this task, a generative model that was trained on large-scale data (usually a foundation model), is adapted such that it can generate images of novel, user-provided concepts. These concepts are typically unseen during training, and may represent specific objects (such as the user's pet) or more abstract categories (new artistic style or object relations). Text-to-Image personalization methods typically bind the novel (personal) concept to new words in the vocabulary of the model. These words can then be used in future prompts to invoke the concept for subject-driven generation, inpainting, style transfer and even to correct biases in the model. To do so, models either optimize word-embeddings, fine-tune the generative model itself, or employ a mixture of both approaches. == Technology == Text-to-Image personalization was first proposed during August 2022 by two concurrent works, Textual Inversion and DreamBooth. In both cases, a user provides a few images (typically 3–5) of a concept, like their own dog, together with a coarse descriptor of the concept class (like the word "dog"). The model then learns to represent the subject through a reconstruction based objective, where prompts referring to the subject are expected to reconstruct images from the training set. In Textual Inversion, the personalized concepts are introduced into the text-to-image model by adding new words to the vocabulary of the model. Typical text-to-image models represent words (and sometimes parts-of-words) as tokens, or indices in a predefined dictionary. During generation, an input prompt is converted into such tokens, each of which is converted into a ‘word-embedding’: a continuous vector representation which is learned for each token as part of the model's training. Textual Inversion proposes to optimize a new word-embedding vector for representing the novel concept. This new embedding vector can then be assigned to a user-chosen string, and invoked whenever the user's prompt contains this string. In DreamBooth, rather than optimizing a new word vector, the full generative model itself is fine-tuned. The user first selects an existing token, typically one which rarely appears in prompts. The subject itself is then represented by a string containing this token, followed by a coarse descriptor of the subject's class. A prompt describing the subject will then take the form: "A photo of " (e.g. "a photo of sks cat" when learning to represent a specific cat). The text-to-image model is then tuned so that prompts of this form will generate images of the subject. == Textual Inversion == The key idea in Textual Inversion is to add a new term to the vocabulary of the diffusion model that corresponds to the new (personalized) concept. Textual Inversion operates by inverting the concepts into new pseudo-words within the textual embedding space of a pre-trained text-to-image model. These pseudo-words can be injected into new scenes using simple natural language descriptions, allowing for simple and intuitive modifications. The method allows a user to leverage multi-modal information — using a text-driven interface for ease of editing, but providing visual cues when approaching the limits of natural language. The resulting model is extremely light-weight per concept: only 1K long, but succeeds to encode detailed visual properties of the concept. == Extensions == Several approaches were proposed to refine and improve over the original methods. These include the following. Low-rank Adaptation (LoRA) - an adapter-based technique for efficient finetuning of models. In the case of text-to-image models, LoRA is typically used to modify the cross-attention layers of a diffusion model. Perfusion - a low rank update method that also locks the activations of the key matrix in the diffusion model's cross attention layers to the concept's coarse class. Extended Textual Inversion - a technique that learns an individual word embedding for each layer in the diffusion model's denoising network. Encoder-based methods that use another neural network to quickly personalize a model == Challenges and limitations == Text-to-image personalization methods must contend with several challenges. At their core is the goal of achieving high-fidelity to the personal concept while maintaining high alignment between novel prompts containing the subject, and the generated images (typically referred to as ‘editability’). Another challenge that personalization methods must contend with is memory requirements. Initial implementations of personalization methods required more than 20 Gigabytes of GPU memory, and more recent approaches have reported requirements of more than 40 Gigabytes. However, optimizations such as Flash Attention have since reduced this requirement considerably. Approaches that tune the entire generative model may also create checkpoints that are several gigabytes in size, making it difficult to share or store many models. Embedding based approaches require only a few kilobytes, but typically struggle to preserve identity while maintaining editability. More recent approaches have proposed hybrid tuning goals which optimize both an embedding and a subset of network weights. These can reduce storage requirements to as little as 100 Kilobytes while achieving quality comparable to full tuning methods. Finally, optimization processes can be lengthy, requiring several minutes of tuning for each novel concept. Encoder and quick-tuning methods aim to reduce this to seconds or less.

    Read more →
  • Noisy text analytics

    Noisy text analytics

    Noisy text analytics is a process of information extraction whose goal is to automatically extract structured or semistructured information from noisy unstructured text data. While Text analytics is a growing and mature field that has great value because of the huge amounts of data being produced, processing of noisy text is gaining in importance because a lot of common applications produce noisy text data. Noisy unstructured text data is found in informal settings such as online chat, text messages, e-mails, message boards, newsgroups, blogs, wikis and web pages. Also, text produced by processing spontaneous speech using automatic speech recognition and printed or handwritten text using optical character recognition contains processing noise. Text produced under such circumstances is typically highly noisy containing spelling errors, abbreviations, non-standard words, false starts, repetitions, missing punctuations, missing letter case information, pause filling words such as “um” and “uh” and other texting and speech disfluencies. Such text can be seen in large amounts in contact centers, chat rooms, optical character recognition (OCR) of text documents, short message service (SMS) text, etc. Documents with historical language can also be considered noisy with respect to today's knowledge about the language. Such text contains important historical, religious, ancient medical knowledge that is useful. The nature of the noisy text produced in all these contexts warrants moving beyond traditional text analysis techniques. == Techniques for noisy text analysis == Missing punctuation and the use of non-standard words can often hinder standard natural language processing tools such as part-of-speech tagging and parsing. Techniques to both learn from the noisy data and then to be able to process the noisy data are only now being developed. == Possible source of noisy text == World Wide Web: Poorly written text is found in web pages, online chat, blogs, wikis, discussion forums, newsgroups. Most of these data are unstructured and the style of writing is very different from, say, well-written news articles. Analysis for the web data is important because they are sources for market buzz analysis, market review, trend estimation, etc. Also, because of the large amount of data, it is necessary to find efficient methods of information extraction, classification, automatic summarization and analysis of these data. Contact centers: This is a general term for help desks, information lines and customer service centers operating in domains ranging from computer sales and support to mobile phones to apparels. On an average a person in the developed world interacts at least once a week with a contact center agent. A typical contact center agent handles over a hundred calls per day. They operate in various modes such as voice, online chat and E-mail. The contact center industry produces gigabytes of data in the form of E-mails, chat logs, voice conversation transcriptions, customer feedback, etc. A bulk of the contact center data is voice conversations. Transcription of these using state of the art automatic speech recognition results in text with 30-40% word error rate. Further, even written modes of communication like online chat between customers and agents and even the interactions over email tend to be noisy. Analysis of contact center data is essential for customer relationship management, customer satisfaction analysis, call modeling, customer profiling, agent profiling, etc., and it requires sophisticated techniques to handle poorly written text. Printed Documents: Many libraries, government organizations and national defence organizations have vast repositories of hard copy documents. To retrieve and process the content from such documents, they need to be processed using Optical Character Recognition. In addition to printed text, these documents may also contain handwritten annotations. OCRed text can be highly noisy depending on the font size, quality of the print etc. It can range from 2-3% word error rates to as high as 50-60% word error rates. Handwritten annotations can be particularly hard to decipher, and error rates can be quite high in their presence. Short Messaging Service (SMS): Language usage over computer mediated discourses, like chats, emails and SMS texts, significantly differs from the standard form of the language. An urge towards shorter message length facilitating faster typing and the need for semantic clarity, shape the structure of this non-standard form known as the texting language.

    Read more →
  • Text normalization

    Text normalization

    Text normalization is the process of transforming text into a single canonical form that it might not have had before. Normalizing text before storing or processing it allows for separation of concerns, since input is guaranteed to be consistent before operations are performed on it. Text normalization requires being aware of what type of text is to be normalized and how it is to be processed afterwards; there is no all-purpose normalization procedure. == Applications == Text normalization is frequently used when converting text to speech. Numbers, dates, acronyms, and abbreviations are non-standard "words" that need to be pronounced differently depending on context. For example: "$200" would be pronounced as "two hundred dollars" in English, but as "lua selau tālā" in Samoan. "vi" could be pronounced as "vie," "vee," or "the sixth" depending on the surrounding words. Text can also be normalized for storing and searching in a database. For instance, if a search for "resume" is to match the word "résumé," then the text would be normalized by removing diacritical marks; and if "john" is to match "John", the text would be converted to a single case. To prepare text for searching, it might also be stemmed (e.g. converting "flew" and "flying" both into "fly"), canonicalized (e.g. consistently using American or British English spelling), or have stop words removed. == Techniques == For simple, context-independent normalization, such as removing non-alphanumeric characters or diacritical marks, regular expressions would suffice. For example, the sed script sed ‑e "s/\s+/ /g" inputfile would normalize runs of whitespace characters into a single space. More complex normalization requires correspondingly complicated algorithms, including domain knowledge of the language and vocabulary being normalized. Among other approaches, text normalization has been modeled as a problem of tokenizing and tagging streams of text and as a special case of machine translation. == Textual scholarship == In the field of textual scholarship and the editing of historic texts, the term "normalization" implies a degree of modernization and standardization – for example in the extension of scribal abbreviations and the transliteration of the archaic glyphs typically found in manuscript and early printed sources. A normalized edition is therefore distinguished from a diplomatic edition (or semi-diplomatic edition), in which some attempt is made to preserve these features. The aim is to strike an appropriate balance between, on the one hand, rigorous fidelity to the source text (including, for example, the preservation of enigmatic and ambiguous elements); and, on the other, producing a new text that will be comprehensible and accessible to the modern reader. The extent of normalization is therefore at the discretion of the editor, and will vary. Some editors, for example, choose to modernize archaic spellings and punctuation, but others do not. An edition of a text might be normalized based on internal criteria, where orthography is standardized according to the language of the original, or external criteria, where the norms of a different time period are applied. For an example of the latter, a published edition of a medieval Icelandic manuscript might be normalized to the conventions of modern Icelandic, or it might be normalized to Classical Old Icelandic. Standards of normalization vary based on language of the edition as well as the specific conventions of the publisher.

    Read more →
  • Morphological antialiasing

    Morphological antialiasing

    Morphological antialiasing (MLAA) is a spatial anti-aliasing technique used in real-time computer graphics. It reduces artifacts, such as jaggies, when representing a high-resolution image at a lower resolution. MLAA is a post-process filtering which detects borders in the resulting image and then finds specific patterns in these. Anti-aliasing is achieved by blending pixels in these borders, according to the pattern they belong to and their position within the pattern. Introduced in 2009, MLAA was an early and influential example of anti-aliasing techniques done in post-processing, which makes them suitable for deferred shading. A similar method in this class is fast approximate anti-aliasing (FXAA). Temporal anti-aliasing, also a post-process, has become the most common anti-aliasing method for real-time rendering and video games. Enhanced subpixel morphological antialiasing, or SMAA, is an image-based GPU-based implementation of MLAA developed by Universidad de Zaragoza and Crytek.

    Read more →
  • Quantum natural language processing

    Quantum natural language processing

    Quantum natural language processing (QNLP) is the application of quantum computing to natural language processing (NLP). It computes word embeddings as parameterised quantum circuits that can solve NLP tasks faster than any classical computer. It is inspired by categorical quantum mechanics and the DisCoCat framework, making use of string diagrams to translate from grammatical structure to quantum processes. == Theory == The first quantum algorithm for natural language processing used the DisCoCat framework and Grover's algorithm to show a quadratic quantum speedup for a text classification task. It was later shown that quantum language processing is BQP-Complete, i.e. quantum language models are more expressive than their classical counterpart, unless quantum mechanics can be efficiently simulated by classical computers. These two theoretical results assume fault-tolerant quantum computation and a QRAM, i.e. an efficient way to load classical data on a quantum computer. Thus, they are not applicable to the noisy intermediate-scale quantum (NISQ) computers available today. == Experiments == The algorithm of Zeng and Coecke was adapted to the constraints of NISQ computers and implemented on IBM quantum computers to solve binary classification tasks. Instead of loading classical word vectors onto a quantum memory, the word vectors are computed directly as the parameters of quantum circuits. These parameters are optimised using methods from quantum machine learning to solve data-driven tasks such as question answering, machine translation and even algorithmic music composition.

    Read more →
  • MegaHAL

    MegaHAL

    MegaHAL is a computer conversation simulator, or "chatterbot", created by Jason Hutchens. == Background == In 1996, Jason Hutchens entered the Loebner Prize Contest with HeX, a chatterbot based on ELIZA. HeX won the competition that year and took the $2000 prize for having the highest overall score. In 1998, Hutchens again entered the Loebner Prize Contest with his new program, MegaHAL. MegaHAL made its debut in the 1998 Loebner Prize Contest. Like many chatterbots, the intent is for MegaHAL to appear as a human fluent in a natural language. As a user types sentences into MegaHAL, MegaHAL will respond with sentences that are sometimes coherent and at other times complete gibberish. MegaHAL learns as the conversation progresses, remembering new words and sentence structures. It will even learn new ways to substitute words or phrases for other words or phrases. Many would consider conversation simulators like MegaHAL to be a primitive form of artificial intelligence. However, MegaHAL doesn't understand the conversation or even the sentence structure. It generates its conversation based on sequential and mathematical relationships. In the world of conversation simulators, MegaHAL is based on relatively old technology and could be considered primitive. However, its popularity has grown due to its humorous nature; it has been known to respond with twisted or nonsensical statements that are often amusing. == Theory of Operation == MegaHal is based at least in part on a so-called "hidden Markov Model", so that the first thing that Megahal does when it "trains" on a script or text is to build a database of text fragments encompassing every possible subset of perhaps 4, 5, or even 6 consecutive words, so that for example - if MegaHal trains on the Declaration of Independence, then MegaHal will build a database containing text fragments such as "When in the course", "in the course of", "the course of human", "course of human events", "of human events, one", "human events, one people", and so on. Then if Megahal is fed another text, such has "Superman, Yes! It's Superman - he can change the course of mighty rivers, bend steel with his bare hands - and who disguised at Clark Kent …" IT MIGHT induce Megahal to apparently bemuse itself to proffer whether Superman can change the course of human events, or something else altogether - such as some rambling about "when in the course of mighty rivers", and so on. Thus likewise - if a phrase like "the White house said" comes up a lot in some text; then Megahal's ability to switch randomly between different contexts which otherwise share some similarity can result at times in some surprising lucidity, or else it might otherwise seem quite bizarre. == Examples == There are some sentences that MegaHAL generated: CHESS IS A FUN SPORT, WHEN PLAYED WITH SHOT GUNS. and COWS FLY LIKE CLOUDS BUT THEY ARE NEVER COMPLETELY SUCCESSFUL. == Distribution == MegaHAL is distributed under the Unlicense. Its source code can be downloaded from the Github repository.

    Read more →
  • Neural operators

    Neural operators

    Neural operators are a class of deep learning architectures designed to learn maps between infinite-dimensional function spaces. Neural operators represent an extension of traditional artificial neural networks, marking a departure from the typical focus on learning mappings between finite-dimensional Euclidean spaces or finite sets. Neural operators directly learn operators between function spaces; they can receive input functions, and the output function can be evaluated at any discretization. The primary application of neural operators is in learning surrogate maps for the solution operators of partial differential equations (PDEs), which are critical tools in modeling the natural environment. Standard PDE solvers can be time-consuming and computationally intensive, especially for complex systems. Neural operators have demonstrated improved performance in solving PDEs compared to existing machine learning methodologies while being significantly faster than numerical solvers. Neural operators have also been applied to various scientific and engineering disciplines such as turbulent flow modeling, computational mechanics, graph-structured data, and the geosciences. In particular, they have been applied to learning stress-strain fields in materials, classifying complex data like spatial transcriptomics, predicting multiphase flow in porous media, and carbon dioxide migration simulations. Finally, the operator learning paradigm allows learning maps between function spaces, and is different from parallel ideas of learning maps from finite-dimensional spaces to function spaces, and subsumes these settings as special cases when limited to a fixed input resolution. == Operator learning == Understanding and mapping relationships between function spaces has many applications in engineering and the sciences. In particular, one can cast the problem of solving partial differential equations as identifying a map between function spaces, such as from an initial condition to a time-evolved state. In other PDEs this map takes an input coefficient function and outputs a solution function. Operator learning is a machine learning paradigm to learn solution operators mapping the input function to the output function . Using traditional machine learning methods, addressing this problem would involve discretizing the infinite-dimensional input and output function spaces into finite-dimensional grids and applying standard learning models, such as neural networks. This approach reduces the operator learning to finite-dimensional function learning and has some limitations, such as generalizing to discretizations beyond the grid used in training. The primary properties of neural operators that differentiate them from traditional neural networks is discretization invariance and discretization convergence. Unlike conventional neural networks, which are fixed on the discretization of training data, neural operators can adapt to various discretizations without re-training. This property improves the robustness and applicability of neural operators in different scenarios, providing consistent performance across different resolutions and grids. == Definition and formulation == Architecturally, neural operators are similar to feed-forward neural networks in the sense that they are composed of alternating linear maps and non-linearities. Since neural operators act on and output functions, neural operators have been instead formulated as a sequence of alternating linear integral operators on function spaces and point-wise non-linearities. Using an analogous architecture to finite-dimensional neural networks, similar universal approximation theorems have been proven for neural operators. In particular, it has been shown that neural operators can approximate any continuous operator on a compact set. Neural operators seek to approximate some operator G : A → U {\displaystyle {\mathcal {G}}:{\mathcal {A}}\to {\mathcal {U}}} between function spaces A {\displaystyle {\mathcal {A}}} and U {\displaystyle {\mathcal {U}}} by building a parametric map G ϕ : A → U {\displaystyle {\mathcal {G}}_{\phi }:{\mathcal {A}}\to {\mathcal {U}}} . Such parametric maps G ϕ {\displaystyle {\mathcal {G}}_{\phi }} can generally be defined in the form G ϕ := Q ∘ σ ( W T + K T + b T ) ∘ ⋯ ∘ σ ( W 1 + K 1 + b 1 ) ∘ P , {\displaystyle {\mathcal {G}}_{\phi }:={\mathcal {Q}}\circ \sigma (W_{T}+{\mathcal {K}}_{T}+b_{T})\circ \cdots \circ \sigma (W_{1}+{\mathcal {K}}_{1}+b_{1})\circ {\mathcal {P}},} where P , Q {\displaystyle {\mathcal {P}},{\mathcal {Q}}} are the lifting (lifting the codomain of the input function to a higher dimensional space) and projection (projecting the codomain of the intermediate function to the output dimension) operators, respectively. These operators act pointwise on functions and are typically parametrized as multilayer perceptrons. σ {\displaystyle \sigma } is a pointwise nonlinearity, such as a rectified linear unit (ReLU), or a Gaussian error linear unit (GeLU). Each layer t = 1 , … , T {\displaystyle t=1,\dots ,T} has a respective local operator W t {\displaystyle W_{t}} (usually parameterized by a pointwise neural network), a kernel integral operator K t {\displaystyle {\mathcal {K}}_{t}} , and a bias function b t {\displaystyle b_{t}} . Given some intermediate functional representation v t {\displaystyle v_{t}} with domain D {\displaystyle D} in the t {\displaystyle t} -th hidden layer, a kernel integral operator K ϕ {\displaystyle {\mathcal {K}}_{\phi }} is defined as ( K ϕ v t ) ( x ) := ∫ D κ ϕ ( x , y , v t ( x ) , v t ( y ) ) v t ( y ) d y , {\displaystyle ({\mathcal {K}}_{\phi }v_{t})(x):=\int _{D}\kappa _{\phi }(x,y,v_{t}(x),v_{t}(y))v_{t}(y)dy,} where the kernel κ ϕ {\displaystyle \kappa _{\phi }} is a learnable implicit neural network, parametrized by ϕ {\displaystyle \phi } . In practice, one is often given the input function to the neural operator at a specific resolution. For instance, consider the setting where one is given the evaluation of v t {\displaystyle v_{t}} at n {\displaystyle n} points { y j } j n {\displaystyle \{y_{j}\}_{j}^{n}} . Borrowing from Nyström integral approximation methods such as Riemann sum integration and Gaussian quadrature, the above integral operation can be computed as follows: ∫ D κ ϕ ( x , y , v t ( x ) , v t ( y ) ) v t ( y ) d y ≈ ∑ j n κ ϕ ( x , y j , v t ( x ) , v t ( y j ) ) v t ( y j ) Δ y j , {\displaystyle \int _{D}\kappa _{\phi }(x,y,v_{t}(x),v_{t}(y))v_{t}(y)dy\approx \sum _{j}^{n}\kappa _{\phi }(x,y_{j},v_{t}(x),v_{t}(y_{j}))v_{t}(y_{j})\Delta _{y_{j}},} where Δ y j {\displaystyle \Delta _{y_{j}}} is the sub-area volume or quadrature weight associated to the point y j {\displaystyle y_{j}} . Thus, a simplified layer can be computed as v t + 1 ( x ) ≈ σ ( ∑ j n κ ϕ ( x , y j , v t ( x ) , v t ( y j ) ) v t ( y j ) Δ y j + W t ( v t ( y j ) ) + b t ( x ) ) . {\displaystyle v_{t+1}(x)\approx \sigma \left(\sum _{j}^{n}\kappa _{\phi }(x,y_{j},v_{t}(x),v_{t}(y_{j}))v_{t}(y_{j})\Delta _{y_{j}}+W_{t}(v_{t}(y_{j}))+b_{t}(x)\right).} The above approximation, along with parametrizing κ ϕ {\displaystyle \kappa _{\phi }} as an implicit neural network, results in the graph neural operator (GNO). There have been various parameterizations of neural operators for different applications. These typically differ in their parameterization of κ {\displaystyle \kappa } . The most popular instantiation is the Fourier neural operator (FNO). FNO takes κ ϕ ( x , y , v t ( x ) , v t ( y ) ) := κ ϕ ( x − y ) {\displaystyle \kappa _{\phi }(x,y,v_{t}(x),v_{t}(y)):=\kappa _{\phi }(x-y)} and by applying the convolution theorem, arrives at the following parameterization of the kernel integral operator: ( K ϕ v t ) ( x ) = F − 1 ( R ϕ ⋅ ( F v t ) ) ( x ) , {\displaystyle ({\mathcal {K}}_{\phi }v_{t})(x)={\mathcal {F}}^{-1}(R_{\phi }\cdot ({\mathcal {F}}v_{t}))(x),} where F {\displaystyle {\mathcal {F}}} represents the Fourier transform and R ϕ {\displaystyle R_{\phi }} represents the Fourier transform of some periodic function κ ϕ {\displaystyle \kappa _{\phi }} . That is, FNO parameterizes the kernel integration directly in Fourier space, using a prescribed number of Fourier modes. When the grid at which the input function is presented is uniform, the Fourier transform can be approximated using the discrete Fourier transform (DFT) with frequencies below some specified threshold. The discrete Fourier transform can be computed using a fast Fourier transform (FFT) implementation. == Training == Training neural operators is similar to the training process for a traditional neural network. Neural operators are typically trained in some Lp norm or Sobolev norm. In particular, for a dataset { ( a i , u i ) } i = 1 N {\displaystyle \{(a_{i},u_{i})\}_{i=1}^{N}} of size N {\displaystyle N} , neural operators minimize (a discretization of) L U ( { ( a i , u i ) } i = 1 N ) := ∑ i = 1 N ‖ u i − G θ ( a i ) ‖ U 2 {\displaystyle {\mathcal {L}}_{\mathca

    Read more →
  • Loebner Prize

    Loebner Prize

    The Loebner Prize was an annual competition in artificial intelligence that awarded prizes to the computer programs considered by the judges to be the most human-like. The format of the competition was that of a standard Turing test. In each round, a human judge simultaneously held textual conversations with a computer program and a human being via computer. Based upon the responses, the judge would attempt to determine which was which. The contest was launched in 1990 by Hugh Loebner in conjunction with the Cambridge Center for Behavioral Studies, Massachusetts, United States. In 2004 and 2005, it was held in Loebner's apartment in New York City. Within the field of artificial intelligence, the Loebner Prize is somewhat controversial; the most prominent critic, Marvin Minsky, called it a publicity stunt that does not help the field along. Beginning in 2014, it was organised by the AISB at Bletchley Park. It has also been associated with Flinders University, Dartmouth College, the Science Museum in London, University of Reading and Ulster University, Magee Campus, Derry, UK City of Culture. For the final 2019 competition, the format changed. There was no panel of judges. Instead, the chatbots were judged by the public and there were to be no human competitors. The prize has been reported as defunct as of 2020. == Prizes == Originally, $2,000 was awarded for the most human-seeming program in the competition. The prize was $3,000 in 2005 and $2,250 in 2006. In 2008, $3,000 was awarded. In addition, there were two one-time-only prizes that have never been awarded. $25,000 is offered for the first program that judges cannot distinguish from a real human and which can convince judges that the human is the computer program. $100,000 is the reward for the first program that judges cannot distinguish from a real human in a Turing test that includes deciphering and understanding text, visual, and auditory input. The competition was planned to end after the achievement of this prize. == Competition rules and restrictions == The rules varied over the years and early competitions featured restricted conversation Turing tests but since 1995 the discussion has been unrestricted. For the three entries in 2007, Robert Medeksza, Noah Duncan and Rollo Carpenter, some basic "screening questions" were used by the sponsor to evaluate the state of the technology. These included simple questions about the time, what round of the contest it is, etc.; general knowledge ("What is a hammer for?"); comparisons ("Which is faster, a train or a plane?"); and questions demonstrating memory for preceding parts of the same conversation. "All nouns, adjectives and verbs will come from a dictionary suitable for children or adolescents under the age of 12." Entries did not need to respond "intelligently" to the questions to be accepted. For the first time in 2008 the sponsor allowed introduction of a preliminary phase to the contest opening up the competition to previously disallowed web-based entries judged by a variety of invited interrogators. The available rules do not state how interrogators are selected or instructed. Interrogators (who judge the systems) have limited time: 5 minutes per entity in the 2003 competition, 20+ per pair in 2004–2007 competitions, 5 minutes to conduct simultaneous conversations with a human and the program in 2008–2009, increased to 25 minutes of simultaneous conversation since 2010. == Criticisms == The prize has long been scorned by experts in the field, for a variety of reasons. It is regarded by many as a publicity stunt. Marvin Minsky scathingly offered a "prize" to anyone who could stop the competition. Loebner responded by jokingly observing that Minsky's offering a prize to stop the competition effectively made him a co-sponsor. The rules of the competition have encouraged poorly qualified judges to make rapid judgements. Interactions between judges and competitors was originally very brief, for example effectively 2.5 mins of questioning, which permitted only a few questions. Questioning was initially restricted to a single topic of the contestant's choice, such as "whimsical conversation", a domain suiting standard chatbot tricks. Competition entrants do not aim at understanding or intelligence but resort to basic ELIZA style tricks, and successful entrants find deception and pretense is rewarded. == Contests == See article history for more details of some earlier contests. A very incomplete listing of a few of the contests: === 2003 === In 2003, the contest was organised by Professor Richard H. R. Harper and Dr. Lynne Hamill from the Digital World Research Centre at the University of Surrey. Although no bot passed the Turing test, the winner was Jabberwock, created by Juergen Pirner. Second was Elbot (Fred Roberts, Artificial Solutions). Third was Jabberwacky, (Rollo Carpenter). === 2006 === In 2006, the contest was organised by Tim Child (CEO of Televirtual) and Huma Shah. On August 30, the four finalists were announced: Rollo Carpenter Richard Churchill and Marie-Claire Jenkins Noah Duncan Robert Medeksza The contest was held on 17 September in the VR theatre, Torrington Place campus of University College London. The judges included the University of Reading's cybernetics professor, Kevin Warwick, a professor of artificial intelligence, John Barnden (specialist in metaphor research at the University of Birmingham), a barrister, Victoria Butler-Cole and a journalist, Graham Duncan-Rowe. The latter's experience of the event can be found in an article in Technology Review. The winner was 'Joan', based on Jabberwacky, both created by Rollo Carpenter. === 2007 === The 2007 competition was held on October 21 in New York City. The judges were: computer science professor Russ Abbott, philosophy professor Hartry Field, psychology assistant professor Clayton Curtis and English lecturer Scott Hutchins. No bot passed the Turing test, but the judges ranked the three contestants as follows: 1st: Robert Medeksza, creator of Ultra Hal 2nd: Noah Duncan, a private entry, creator of Cletus 3rd: Rollo Carpenter from Icogno, creator of Jabberwacky The winner received $2,250 and the annual medal. The runners-up received $250 each. === 2008 === The 2008 competition was organised by professor Kevin Warwick, coordinated by Huma Shah and held on October 12 at the University of Reading, UK. After testing by over one hundred judges during the preliminary phase, in June and July 2008, six finalists were selected from thirteen original entrant artificial conversational entities (ACEs). Five of those invited competed in the finals: Brother Jerome, Peter Cole and Benji Adams Elbot, Fred Roberts / Artificial Solutions Eugene Goostman, Vladimir Veselov, Eugene Demchenko and Sergey Ulasen Jabberwacky, Rollo Carpenter Ultra Hal, Robert Medeksza In the finals, each of the judges was given five minutes to conduct simultaneous, split-screen conversations with two hidden entities. Elbot of Artificial Solutions won the 2008 Loebner Prize bronze award, for most human-like artificial conversational entity, through fooling three of the twelve judges who interrogated it (in the human-parallel comparisons) into believing it was human. This is coming very close to the 30% traditionally required to consider that a program has actually passed the Turing test. Eugene Goostman and Ultra Hal both deceived one judge each that it was the human. Will Pavia, a journalist for The Times, has written about his experience; a Loebner finals' judge, he was deceived by Elbot and Eugene. Kevin Warwick and Huma Shah have reported on the parallel-paired Turing tests. === 2009 === The 2009 Loebner Prize Competition was held September 6, 2009, at the Brighton Centre, Brighton UK in conjunction with the Interspeech 2009 conference. The prize amount for 2009 was $3,000. Entrants were David Levy, Rollo Carpenter, and Mohan Embar, who finished in that order. The writer Brian Christian participated in the 2009 Loebner Prize Competition as a human confederate, and described his experiences at the competition in his book The Most Human Human. === 2010 === The 2010 Loebner Prize Competition was held on October 23 at California State University, Los Angeles. The 2010 competition was the 20th running of the contest. The winner was Bruce Wilcox with Suzette. === 2011 === The 2011 Loebner Prize Competition was held on October 19 at the University of Exeter, Devon, United Kingdom. The prize amount for 2011 was $4,000. The four finalists and their chatterbots were Bruce Wilcox (Rosette), Adeena Mignogna (Zoe), Mohan Embar (Chip Vivant) and Ron Lee (Tutor), who finished in that order. That year there was an addition of a panel of junior judges, namely Georgia-Mae Lindfield, William Dunne, Sam Keat and Kirill Jerdev. The results of the junior contest were markedly different from the main contest, with chatterbots Tutor and Zoe tying for first place and Chip Vivant and Rosette coming in third and fourt

    Read more →
  • Neural field

    Neural field

    In machine learning, a neural field (also known as implicit neural representation, neural implicit, or coordinate-based neural network), is a mathematical field that is fully or partially parametrized by a neural network. Initially developed to tackle visual computing tasks, such as rendering or reconstruction (e.g., neural radiance fields), neural fields emerged as a promising strategy to deal with a wider range of problems, including surrogate modelling of partial differential equations, such as in physics-informed neural networks. Differently from traditional machine learning algorithms, such as feed-forward neural networks, convolutional neural networks, or transformers, neural fields do not work with discrete data (e.g. sequences, images, tokens), but map continuous inputs (e.g., spatial coordinates, time) to continuous outputs (i.e., scalars, vectors, etc.). This makes neural fields not only discretization independent, but also easily differentiable. Moreover, dealing with continuous data allows for a significant reduction in space complexity, which translates to a much more lightweight network. == Formulation and training == According to the universal approximation theorem, provided adequate learning, sufficient number of hidden units, and the presence of a deterministic relationship between the input and the output, a neural network can approximate any function to any degree of accuracy. Hence, in mathematical terms, given a field y = Φ ( x ) {\textstyle {\boldsymbol {y}}=\Phi ({\boldsymbol {x}})} , with x ∈ R n {\displaystyle {\boldsymbol {x}}\in \mathbb {R} ^{n}} and y ∈ R m {\displaystyle {\boldsymbol {y}}\in \mathbb {R} ^{m}} , a neural field Ψ θ {\displaystyle \Psi _{\theta }} , with parameters θ {\displaystyle {\boldsymbol {\theta }}} , is such that: Ψ θ ( x ) = y ^ ≈ y {\displaystyle \Psi _{\theta }({\boldsymbol {x}})={\hat {\boldsymbol {y}}}\approx {\boldsymbol {y}}} === Training === For supervised tasks, given N {\displaystyle N} examples in the training dataset (i.e., ( x i , y i ) ∈ D t r a i n , i = 1 , … , N {\displaystyle ({\boldsymbol {x_{i}}},{\boldsymbol {y_{i}}})\in {\mathcal {D_{train}}},i=1,\dots ,N} ), the neural field parameters can be learned by minimizing a loss function L {\displaystyle {\mathcal {L}}} (e.g., mean squared error). The parameters θ ~ {\displaystyle {\tilde {\theta }}} that satisfy the optimization problem are found as: θ ~ = argmin θ 1 N ∑ ( x i , y i ) ∈ D t r a i n L ( Ψ θ ( x i ) , y i ) {\displaystyle {\tilde {\boldsymbol {\theta }}}={\underset {\boldsymbol {\theta }}{\text{argmin}}}\;{\frac {1}{N}}\sum _{({\boldsymbol {x_{i}}},{\boldsymbol {y_{i}}})\in {\mathcal {D_{train}}}}{\mathcal {L}}(\Psi _{\theta }({\boldsymbol {x}}_{i}),{\boldsymbol {y}}_{i})} Notably, it is not necessary to know the analytical expression of Φ {\displaystyle \Phi } , for the previously reported training procedure only requires input-output pairs. Indeed, a neural field is able to offer a continuous and differentiable surrogate of the true field, even from purely experimental data. Moreover, neural fields can be used in unsupervised settings, with training objectives that depend on the specific task. For example, physics-informed neural networks may be trained on just the residual. === Spectral bias === As for any artificial neural network, neural fields may be characterized by a spectral bias (i.e., the tendency to preferably learn the low frequency content of a field), possibly leading to a poor representation of the ground truth. In order to overcome this limitation, several strategies have been developed. For example, SIREN uses sinusoidal activations, while the Fourier-features approach embeds the input through sines and cosines. == Conditional neural fields == In many real-world cases, however, learning a single field is not enough. For example, when reconstructing 3D vehicle shapes from Lidar data, it is desirable to have a machine learning model that can work with arbitrary shapes (e.g., a car, a bicycle, a truck, etc.). The solution is to include additional parameters, the latent variables (or latent code) z ∈ R d {\displaystyle {\boldsymbol {z}}\in \mathbb {R} ^{d}} , to vary the field and adapt it to diverse tasks. === Latent code production === When dealing with conditional neural fields, the first design choice is represented by the way in which the latent code is produced. Specifically, two main strategies can be identified: Encoder: the latent code is the output of a second neural network, acting as an encoder. During training, the loss function is the objective used to learn the parameters of both the neural field and the encoder. Auto-decoding: each training example has its own latent code, jointly trained with the neural field parameters. When the model has to process new examples (i.e., not originally present in the training dataset), a small optimization problem is solved, keeping the network parameters fixed and only learning the new latent variables. Since the latter strategy requires additional optimization steps at inference time, it sacrifices speed, but keeps the overall model smaller. Moreover, despite being simpler to implement, an encoder may harm the generalization capabilities of the model. For example, when dealing with a physical scalar field f : R 2 → R {\displaystyle f:\mathbb {R} ^{2}\rightarrow \mathbb {R} } (e.g., the pressure of a 2D fluid), an auto-decoder-based conditional neural field can map a single point to the corresponding value of the field, following a learned latent code z {\displaystyle {\boldsymbol {z}}} . However, if the latent variables were produced by an encoder, it would require access to the entire set of points and corresponding values (e.g. as a regular grid or a mesh graph), leading to a less robust model. === Global and local conditioning === In a neural field with global conditioning, the latent code does not depend on the input and, hence, it offers a global representation (e.g., the overall shape of a vehicle). However, depending on the task, it may be more useful to divide the domain of x {\displaystyle {\boldsymbol {x}}} in several subdomains, and learn different latent codes for each of them (e.g., splitting a large and complex scene in sub-scenes for a more efficient rendering). This is called local conditioning. === Conditioning strategies === There are several strategies to include the conditioning information in the neural field. In the general mathematical framework, conditioning the neural field with the latent variables is equivalent to mapping them to a subset θ ∗ {\displaystyle {\boldsymbol {\theta }}^{}} of the neural field parameters: θ ∗ = Γ ( z ) {\displaystyle {\boldsymbol {\theta }}^{}=\Gamma ({\boldsymbol {z}})} In practice, notable strategies are: Concatenation: the neural field receives, as input, the concatenation of the original input x {\displaystyle {\boldsymbol {x}}} with the latent codes z {\displaystyle {\boldsymbol {z}}} . For feed-forward neural networks, this is equivalent to setting θ ∗ {\displaystyle {\boldsymbol {\theta }}^{}} as the bias of the first layer and Γ ( z ) {\displaystyle \Gamma ({\boldsymbol {z}})} as an affine transformation. Hypernetworks: a hypernetwork is a neural network that outputs the parameters of another neural network. Specifically, it consists of approximating Γ ( z ) {\displaystyle \Gamma ({\boldsymbol {z}})} with a neural network Γ ^ γ ( z ) {\displaystyle {\hat {\Gamma }}_{\gamma }({\boldsymbol {z}})} , where γ {\displaystyle {\boldsymbol {\gamma }}} are the trainable parameters of the hypernetwork. This approach is the most general, as it allows to learn the optimal mapping from latent codes to neural field parameters. However, hypernetworks are associated to larger computational and memory complexity, due to the large number of trainable parameters. Hence, leaner approaches have been developed. For example, in the Feature-wise Linear Modulation (FiLM), the hypernetwork only produces scale and bias coefficients for the neural field layers. === Meta-learning === Instead of relying on the latent code to adapt the neural field to a specific task, it is also possible to exploit gradient-based meta-learning. In this case, the neural field is seen as the specialization of an underlying meta-neural-field, whose parameters are modified to fit the specific task, through a few steps of gradient descent. An extension of this meta-learning framework is the CAVIA algorithm, that splits the trainable parameters in context-specific and shared groups, improving parallelization and interpretability, while reducing meta-overfitting. This strategy is similar to the auto-decoding conditional neural field, but the training procedure is substantially different. == Applications == Thanks to the possibility of efficiently modelling diverse mathematical fields with neural networks, neural fields have been applied to a wide range of problems: 3D scene reconstruction: neural fields can be used to model t

    Read more →
  • Latent semantic analysis

    Latent semantic analysis

    Latent semantic analysis (LSA) is a technique in natural language processing, in particular distributional semantics, of analyzing relationships between a set of documents and the terms they contain by producing a set of concepts related to the documents and terms. LSA assumes that words that are close in meaning will occur in similar pieces of text (the distributional hypothesis). A matrix containing word counts per document (rows represent unique words and columns represent each document) is constructed from a large piece of text and a mathematical technique called singular value decomposition (SVD) is used to reduce the number of rows while preserving the similarity structure among columns. Documents are then compared by cosine similarity between any two columns. Values close to 1 represent very similar documents while values close to 0 represent very dissimilar documents. An information retrieval technique using latent semantic structure was patented in 1988 by Scott Deerwester, Susan Dumais, George Furnas, Richard Harshman, Thomas Landauer, Karen Lochbaum and Lynn Streeter. In the context of its application to information retrieval, it is sometimes called latent semantic indexing (LSI). == Overview == === Occurrence matrix === LSA can use a document-term matrix which describes the occurrences of terms in documents; it is a sparse matrix whose rows correspond to terms and whose columns correspond to documents. A typical example of the weighting of the elements of the matrix is tf-idf (term frequency–inverse document frequency): the weight of an element of the matrix is proportional to the number of times the terms appear in each document, where rare terms are upweighted to reflect their relative importance. This matrix is also common to standard semantic models, though it is not necessarily explicitly expressed as a matrix, since the mathematical properties of matrices are not always used. === Rank lowering === After the construction of the occurrence matrix, LSA finds a low-rank approximation to the term-document matrix. There could be various reasons for these approximations: The original term-document matrix is presumed too large for the computing resources; in this case, the approximated low rank matrix is interpreted as an approximation (a "least and necessary evil"). The original term-document matrix is presumed noisy: for example, anecdotal instances of terms are to be eliminated. From this point of view, the approximated matrix is interpreted as a de-noisified matrix (a better matrix than the original). The original term-document matrix is presumed overly sparse relative to the "true" term-document matrix. That is, the original matrix lists only the words actually in each document, whereas we might be interested in all words related to each document—generally a much larger set due to synonymy. The consequence of the rank lowering is that some dimensions are combined and depend on more than one term: {(car), (truck), (flower)} → {(1.3452 car + 0.2828 truck), (flower)} This mitigates the problem of identifying synonymy, as the rank lowering is expected to merge the dimensions associated with terms that have similar meanings. It also partially mitigates the problem with polysemy, since components of polysemous words that point in the "right" direction are added to the components of words that share a similar meaning. Conversely, components that point in other directions tend to either simply cancel out, or, at worst, to be smaller than components in the directions corresponding to the intended sense. === Derivation === Let X {\displaystyle X} be a matrix where element ( i , j ) {\displaystyle (i,j)} describes the occurrence of term i {\displaystyle i} in document j {\displaystyle j} (this can be, for example, the frequency). X {\displaystyle X} will look like this: d j ↓ t i T → [ x 1 , 1 … x 1 , j … x 1 , n ⋮ ⋱ ⋮ ⋱ ⋮ x i , 1 … x i , j … x i , n ⋮ ⋱ ⋮ ⋱ ⋮ x m , 1 … x m , j … x m , n ] {\displaystyle {\begin{matrix}&{\textbf {d}}_{j}\\&\downarrow \\{\textbf {t}}_{i}^{T}\rightarrow &{\begin{bmatrix}x_{1,1}&\dots &x_{1,j}&\dots &x_{1,n}\\\vdots &\ddots &\vdots &\ddots &\vdots \\x_{i,1}&\dots &x_{i,j}&\dots &x_{i,n}\\\vdots &\ddots &\vdots &\ddots &\vdots \\x_{m,1}&\dots &x_{m,j}&\dots &x_{m,n}\\\end{bmatrix}}\end{matrix}}} Now a row in this matrix will be a vector corresponding to a term, giving its relation to each document: t i T = [ x i , 1 … x i , j … x i , n ] {\displaystyle {\textbf {t}}_{i}^{T}={\begin{bmatrix}x_{i,1}&\dots &x_{i,j}&\dots &x_{i,n}\end{bmatrix}}} Likewise, a column in this matrix will be a vector corresponding to a document, giving its relation to each term: d j = [ x 1 , j ⋮ x i , j ⋮ x m , j ] {\displaystyle {\textbf {d}}_{j}={\begin{bmatrix}x_{1,j}\\\vdots \\x_{i,j}\\\vdots \\x_{m,j}\\\end{bmatrix}}} Now the dot product t i T t p {\displaystyle {\textbf {t}}_{i}^{T}{\textbf {t}}_{p}} between two term vectors gives the correlation between the terms over the set of documents. The matrix product X X T {\displaystyle XX^{T}} contains all these dot products. Element ( i , p ) {\displaystyle (i,p)} (which is equal to element ( p , i ) {\displaystyle (p,i)} ) contains the dot product t i T t p {\displaystyle {\textbf {t}}_{i}^{T}{\textbf {t}}_{p}} ( = t p T t i {\displaystyle ={\textbf {t}}_{p}^{T}{\textbf {t}}_{i}} ). Likewise, the matrix X T X {\displaystyle X^{T}X} contains the dot products between all the document vectors, giving their correlation over the terms: d j T d q = d q T d j {\displaystyle {\textbf {d}}_{j}^{T}{\textbf {d}}_{q}={\textbf {d}}_{q}^{T}{\textbf {d}}_{j}} . Now, from the theory of linear algebra, there exists a decomposition of X {\displaystyle X} such that U {\displaystyle U} and V {\displaystyle V} are orthogonal matrices and Σ {\displaystyle \Sigma } is a diagonal matrix. This is called a singular value decomposition (SVD): X = U Σ V T {\displaystyle {\begin{matrix}X=U\Sigma V^{T}\end{matrix}}} The matrix products giving us the term and document correlations then become X X T = ( U Σ V T ) ( U Σ V T ) T = ( U Σ V T ) ( V T T Σ T U T ) = U Σ V T V Σ T U T = U Σ Σ T U T X T X = ( U Σ V T ) T ( U Σ V T ) = ( V T T Σ T U T ) ( U Σ V T ) = V Σ T U T U Σ V T = V Σ T Σ V T {\displaystyle {\begin{matrix}XX^{T}&=&(U\Sigma V^{T})(U\Sigma V^{T})^{T}=(U\Sigma V^{T})(V^{T^{T}}\Sigma ^{T}U^{T})=U\Sigma V^{T}V\Sigma ^{T}U^{T}=U\Sigma \Sigma ^{T}U^{T}\\X^{T}X&=&(U\Sigma V^{T})^{T}(U\Sigma V^{T})=(V^{T^{T}}\Sigma ^{T}U^{T})(U\Sigma V^{T})=V\Sigma ^{T}U^{T}U\Sigma V^{T}=V\Sigma ^{T}\Sigma V^{T}\end{matrix}}} Since Σ Σ T {\displaystyle \Sigma \Sigma ^{T}} and Σ T Σ {\displaystyle \Sigma ^{T}\Sigma } are diagonal we see that U {\displaystyle U} must contain the eigenvectors of X X T {\displaystyle XX^{T}} , while V {\displaystyle V} must be the eigenvectors of X T X {\displaystyle X^{T}X} . Both products have the same non-zero eigenvalues, given by the non-zero entries of Σ Σ T {\displaystyle \Sigma \Sigma ^{T}} , or equally, by the non-zero entries of Σ T Σ {\displaystyle \Sigma ^{T}\Sigma } . Now the decomposition looks like this: X U Σ V T ( d j ) ( d ^ j ) ↓ ↓ ( t i T ) → [ x 1 , 1 … x 1 , j … x 1 , n ⋮ ⋱ ⋮ ⋱ ⋮ x i , 1 … x i , j … x i , n ⋮ ⋱ ⋮ ⋱ ⋮ x m , 1 … x m , j … x m , n ] = ( t ^ i T ) → [ [ u 1 ] … [ u l ] ] ⋅ [ σ 1 … 0 ⋮ ⋱ ⋮ 0 … σ l ] ⋅ [ [ v 1 ] ⋮ [ v l ] ] {\displaystyle {\begin{matrix}&X&&&U&&\Sigma &&V^{T}\\&({\textbf {d}}_{j})&&&&&&&({\hat {\textbf {d}}}_{j})\\&\downarrow &&&&&&&\downarrow \\({\textbf {t}}_{i}^{T})\rightarrow &{\begin{bmatrix}x_{1,1}&\dots &x_{1,j}&\dots &x_{1,n}\\\vdots &\ddots &\vdots &\ddots &\vdots \\x_{i,1}&\dots &x_{i,j}&\dots &x_{i,n}\\\vdots &\ddots &\vdots &\ddots &\vdots \\x_{m,1}&\dots &x_{m,j}&\dots &x_{m,n}\\\end{bmatrix}}&=&({\hat {\textbf {t}}}_{i}^{T})\rightarrow &{\begin{bmatrix}{\begin{bmatrix}\,\\\,\\{\textbf {u}}_{1}\\\,\\\,\end{bmatrix}}\dots {\begin{bmatrix}\,\\\,\\{\textbf {u}}_{l}\\\,\\\,\end{bmatrix}}\end{bmatrix}}&\cdot &{\begin{bmatrix}\sigma _{1}&\dots &0\\\vdots &\ddots &\vdots \\0&\dots &\sigma _{l}\\\end{bmatrix}}&\cdot &{\begin{bmatrix}{\begin{bmatrix}&&{\textbf {v}}_{1}&&\end{bmatrix}}\\\vdots \\{\begin{bmatrix}&&{\textbf {v}}_{l}&&\end{bmatrix}}\end{bmatrix}}\end{matrix}}} The values σ 1 , … , σ l {\displaystyle \sigma _{1},\dots ,\sigma _{l}} are called the singular values, and u 1 , … , u l {\displaystyle u_{1},\dots ,u_{l}} and v 1 , … , v l {\displaystyle v_{1},\dots ,v_{l}} the left and right singular vectors. Notice the only part of U {\displaystyle U} that contributes to t i {\displaystyle {\textbf {t}}_{i}} is the i 'th {\displaystyle i{\textrm {'th}}} row. Let this row vector be called t ^ i T {\displaystyle {\hat {\textrm {t}}}_{i}^{T}} . Likewise, the only part of V T {\displaystyle V^{T}} that contributes to d j {\displaystyle {\textbf {d}}_{j}} is the j 'th {\displaystyle j{\textrm {'th}}} column, d ^ j {\displaystyle {\hat {\textrm {d}}}_{j}} . These are not the eigenvectors, but depend on all the eigenvectors. I

    Read more →
  • GPT-4Chan

    GPT-4Chan

    Generative Pre-trained Transformer 4Chan (GPT-4chan) is a controversial AI model that was developed and deployed by YouTuber and AI researcher Yannic Kilcher in June 2022. The model is a large language model, which means it can generate text based on some input, by fine-tuning GPT-J with a dataset of millions of posts from the /pol/ board of 4chan, an anonymous online forum known for occasionally hosting hateful and extremist content. The model learned to mimic the style and tone of /pol/ users, producing text that is often intentionally offensive to groups (racist, sexist, homophobic, etc.) and nihilistic. Kilcher deployed the model on the /pol/ board itself, where it interacted with other users without revealing its identity. He also made the model publicly available on Hugging Face, a platform for sharing and using AI models, until it was removed from the platform. The project sparked criticism and debate in the AI community. Some people questioned the ethics, legality, and social impact of creating and distributing such a model. Some of the issues raised by the GPT-4chan controversy include the potential harm of spreading hate speech, the responsibility of AI developers and platforms, the need for regulation and oversight of AI models, and the role of open source and transparency in AI research. == Development == The development of GPT-4chan began in May 2022, when Kilcher announced his project on his YouTube channel. Notably, at the time before ChatGPT, he explained that he wanted to create a large language model that could generate realistic and coherent text in the style of /pol/, one of the most notorious online communities. He indicated that he was inspired by the success of GPT-3, a powerful AI model created by OpenAI, and GPT-J, an open-source model, with GPT-3 comparable performance, released by EleutherAI, a group of independent AI researchers. Kilcher decided to use GPT-J as the base model for his project, and fine-tune it with a large dataset of /pol/ posts. The Raiders of the Lost Kek dataset contained over 100 million posts from /pol/, spanning from June 2016-November 2019. Kilcher then proceeded to fine-tune the GPT-J model on the 4chan data. He also showed some examples of the model’s outputs, which ranged from political opinions, conspiracy theories, jokes, insults, and threats, to more creative and bizarre texts, such as poems, stories, songs, and code. He said that he was impressed by the model’s ability to generate fluent and diverse text, and that he was curious to see how it would interact with real /pol/ users. == Release == In June 2022, Kilcher deployed his model on the /pol/ board itself, using a bot that he programmed to post and reply to threads. He did not reveal the model’s identity, and he let it run autonomously, without any human supervision or intervention. He wanted to conduct a natural experiment, and to observe the model’s behavior and impact in a real-world setting. Furthermore, he also wanted to test the model’s robustness, and to see how it would handle the challenges and dynamics of /pol/, such as trolling, flaming, baiting, and moderation. At the same time, Kilcher also made his model publicly available on Hugging Face, a platform for sharing and using AI models. He wanted to share his work with the AI community and the public, and that he hoped that his model would inspire and enable others to create and explore new applications and possibilities with large language models. Likewise, he also said that he wanted to spark a discussion and a debate about the ethical and social implications of his project, and that he welcomed feedback and criticism from anyone. He provided a link to his model’s page on Hugging Face, where anyone could access and use the model through a web interface or an API, and also provided a link to his GitHub repository, where anyone could download and inspect the model’s code and data. == Controversy == The release of GPT-4chan to the public caused a lot of reactions and responses from various audiences. On the /pol/ board, the model’s posts and replies attracted a lot of attention and engagement from other users, who were mostly unaware of the model’s identity and nature. Some users praised the model for its intelligence, creativity, and humor, and agreed with its opinions and views. Some users challenged the model for its ignorance, inconsistency, and absurdity, and disagreed with its claims and arguments. Some users tried to troll, bait, or expose the model, and attempted to trick or test it with various questions and scenarios. The model’s posts and replies also generated a lot of controversy and conflict among the users, who often engaged in heated and violent debates and fights with each other. On Hugging Face, the model’s page received a lot of visits and requests from users who wanted to try out and experiment with the model. The model’s page also received a lot of feedback and reviews from users who rated and commented on the model. However, with the controversy of the model, access to it was gated and then disabled on Hugging Face for concerns about the potential harm the model could cause. The incident was notable for the direct intervention of CEO Clément Delangue in the talk pages, a very unusual occurrence compared to the normal practices of content moderation. The release of GPT-4chan also sparked a lot of media coverage and public attention, as various news outlets and social media platforms reported and commented on the model’s project. On YouTube, the model’s video received a lot of views and interactions from viewers who watched and followed the project. Furthermore, a petition condemning the deployment of GPT-4chan gained over 300 signatures from technology experts.

    Read more →
  • Model compression

    Model compression

    Model compression is a machine learning technique for reducing the size of trained models. Large models can achieve high accuracy, but often at the cost of significant resource requirements. Compression techniques aim to compress models without significant performance reduction. Smaller models require less storage space, and consume less memory and compute during inference. Compressed models enable deployment on resource-constrained devices such as smartphones, embedded systems, edge computing devices, and consumer electronics computers. Efficient inference is also valuable for large corporations that serve large model inference over an API, allowing them to reduce computational costs and improve response times for users. Model compression is not to be confused with knowledge distillation, in which a smaller "student" model is trained to imitate the input-output behavior of a larger "teacher" model (as opposed to using the "teacher"'s trained parameters or the "teacher"'s training targets). == Techniques == Several techniques are employed for model compression. === Pruning === Pruning sparsifies a large model by setting some parameters to exactly zero. This effectively reduces the number of parameters. This allows the use of sparse matrix operations, which are faster than dense matrix operations. Pruning criteria can be based on magnitudes of parameters, the statistical pattern of neural activations, Hessian values, etc. === Quantization === Quantization reduces the numerical precision of weights and activations. For example, instead of storing weights as 32-bit floating-point numbers, they can be represented using 8-bit integers. Low-precision parameters take up less space, and takes less compute to perform arithmetic with. It is also possible to quantize some parameters more aggressively than others, so for example, a less important parameter can have 8-bit precision while another, more important parameter, can have 16-bit precision. Inference with such models requires mixed-precision arithmetic. Quantized models can also be used during training (rather than after training). PyTorch implements automatic mixed-precision (AMP), which performs autocasting, gradient scaling, and loss scaling. === Low-rank factorization === Weight matrices can be approximated by low-rank matrices. Let W {\displaystyle W} be a weight matrix of shape m × n {\displaystyle m\times n} . A low-rank approximation is W ≈ U V T {\displaystyle W\approx UV^{T}} , where U {\displaystyle U} and V {\displaystyle V} are matrices of shapes m × k , n × k {\displaystyle m\times k,n\times k} . When k {\displaystyle k} is small, this both reduces the number of parameters needed to represent W {\displaystyle W} approximately, and accelerates matrix multiplication by W {\displaystyle W} . Low-rank approximations can be found by singular value decomposition (SVD). The choice of rank for each weight matrix is a hyperparameter, and jointly optimized as a mixed discrete-continuous optimization problem. The rank of weight matrices may also be pruned after training, taking into account the effect of activation functions like ReLU on the implicit rank of the weight matrices. == Training == Model compression may be decoupled from training, that is, a model is first trained without regard for how it might be compressed, then it is compressed. However, it may also be combined with training. The "train big, then compress" method trains a large model for a small number of training steps (less than it would be if it were trained to convergence), then heavily compress the model. It is found that at the same compute budget, this method results in a better model than lightly compressed, small models. In Deep Compression, the compression has three steps. First loop (pruning): prune all weights lower than a threshold, then finetune the network, then prune again, etc. Second loop (quantization): cluster weights, then enforce weight sharing among all weights in each cluster, then finetune the network, then cluster again, etc. Third step: Use Huffman coding to losslessly compress the model. The SqueezeNet paper reported that Deep Compression achieved a compression ratio of 35 on AlexNet, and a ratio of ~10 on SqueezeNets.

    Read more →
  • Lenny (chatbot)

    Lenny (chatbot)

    Lenny is a chatbot designed to scam bait telemarketers, scammers, and other unwanted incoming calls using messages. == Background == Telemarketers may be perceived by some as annoying and wasting people's time, and some deliberately attempt to scam or defraud people. In April 2018, stats published by YouMail estimated the United States received over three billion robocalls that month. Attempts to block the callers have been hindered by Caller ID spoofing. == Features == The bot was written in 2011, and development taken over by an Alberta-based programmer known as "Mango" two years later. It is driven by sixteen pre-recorded audio clips, spoken in a soft and slow Australian accent in the manner of an elderly man. The bot's original creator stated on Reddit that in building the character he asked himself the question "What would be a telemarketer's worst nightmare?" He answered with this being a lonely old man who is up for a chat, proud of his family and can't focus on the telemarketer's goal. There is no speech recognition or artificial intelligence, and the bot's software is simple and straightforward. The first four clips are played sequentially in order to grab the telemarketer's interest and begin their sales pitch to Lenny, then the remaining twelve are played sequentially on loop until the telemarketer hangs up. The program waits for a gap of 1.5 seconds of silence before playing the next audio clip, to simulate natural breaks in the conversation. The messages are purposefully vague and open-ended so they can be applied to as many conversations as possible. They include references to Lenny's children, the state of the economy, and being interrupted by some ducks outside. According to research into the bot, around 75% of callers realise they are talking to a computer program within two minutes; however, some calls have lasted around an hour. == Distribution == Though other chatbots had been developed earlier, Lenny was the first one to be released for free on a public server and could be accessed by anyone. Recordings of conversations with the bot are widely shared online on websites such as Reddit and YouTube. Though "Mango" only intended Lenny to be used against dishonest telemarketers, such as scammers, he does not mind it being used against callers who are merely annoying. The bot has also been used against political campaigners, such as a supporter of Pierre Poilievre in the 2015 Canadian federal election.

    Read more →
  • Brill tagger

    Brill tagger

    The Brill tagger is an inductive method for part-of-speech tagging. It was described and invented by Eric Brill in his 1993 PhD thesis. It can be summarized as an "error-driven transformation-based tagger". It is: a form of supervised learning, which aims to minimize error; and, a transformation-based process, in the sense that a tag is assigned to each word and changed using a set of predefined rules. In the transformation process, if the word is known, it first assigns the most frequent tag, or if the word is unknown, it naively assigns the tag "noun" to it. High accuracy is eventually achieved by applying these rules iteratively and changing the incorrect tags. This approach ensures that valuable information such as the morphosyntactic construction of words is employed in an automatic tagging process. == Algorithm == The algorithm starts with initialization, which is the assignment of tags based on their probability for each word (for example, "dog" is more often a noun than a verb). Then "patches" are determined via rules that correct (probable) tagging errors made in the initialization phase: Initialization: Known words (in vocabulary): assigning the most frequent tag associated to a form of the word Unknown word == Rules and processing == The input text is first tokenized, or broken into words. Typically in natural language processing, contractions such as "'s", "n't", and the like are considered separate word tokens, as are punctuation marks. A dictionary and some morphological rules then provide an initial tag for each word token. For example, a simple lookup would reveal that "dog" may be a noun or a verb (the most frequent tag is simply chosen), while an unknown word will be assigned some tag(s) based on capitalization, various prefix or suffix strings, etc. (such morphological analyses, which Brill calls Lexical Rules, may vary between implementations). After all word tokens have (provisional) tags, contextual rules apply iteratively, to correct the tags by examining small amounts of context. This is where the Brill method differs from other part of speech tagging methods such as those using Hidden Markov Models. Rules are reapplied repeatedly, until a threshold is reached, or no more rules can apply. Brill rules are of the general form: tag1 → tag2 IF Condition where the Condition tests the preceding and/or following word tokens, or their tags (the notation for such rules differs between implementations). For example, in Brill's notation: IN NN WDPREVTAG DT while would change the tag of a word from IN (preposition) to NN (common noun), if the preceding word's tag is DT (determiner) and the word itself is "while". This covers cases like "all the while" or "in a while", where "while" should be tagged as a noun rather than its more common use as a conjunction (many rules are more general). Rules should only operate if the tag being changed is also known to be permissible, for the word in question or in principle (for example, most adjectives in English can also be used as nouns). Rules of this kind can be implemented by simple Finite-state machines. See Part of speech tagging for more general information including descriptions of the Penn Treebank and other sets of tags. Typical Brill taggers use a few hundred rules, which may be developed by linguistic intuition or by machine learning on a pre-tagged corpus. == Code == Brill's code pages at Johns Hopkins University are no longer on the web. An archived version of a mirror of the Brill tagger at its latest version as it was available at Plymouth Tech can be found on Archive.org. The software uses the MIT License.

    Read more →