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.
Tertiary review
In software engineering, a tertiary review is a systematic review of systematic reviews. It is also referred to as a tertiary study in the software engineering literature. However, Umbrella review is the term more commonly used in medicine. Kitchenham et al. suggest that methodologically there is no difference between a systematic review and a tertiary review. However, as the software engineering community has started performing tertiary reviews new concerns unique to tertiary reviews have surfaced. These include the challenge of quality assessment of systematic reviews, search validation and the additional risk of double counting. == Examples of Tertiary reviews in software engineering literature == Test quality Machine Learning Test-driven development
AI Futures Project
The AI Futures Project is a nonprofit research organization based in the United States that specializes in forecasting the development and societal impact of advanced artificial intelligence. The organization is best known for its 2025 scenario forecast, AI 2027, which examines the potential near-term emergence of artificial general intelligence (AGI) and its possible global consequences. == History == The AI Futures Project was founded in 2025 by Daniel Kokotajlo, a former researcher in the governance division of OpenAI. Kokotajlo resigned from OpenAI in April 2024, expressing concerns that the company prioritized rapid product development over AI safety and was advancing without sufficient safeguards. He founded the nonprofit to conduct independent forecasting and policy research. The organization is registered as a 501(c)(3) nonprofit in the United States and is funded through donations. It operates with a small research staff and network of advisors drawn from fields including AI policy, forecasting, and risk analysis. == Activities == The mission of the AI Futures Project is to develop detailed scenario forecasts of the trajectory of advanced AI systems to inform policymakers, researchers, and the public. In addition to written reports, the group has conducted tabletop exercises and workshops based on its scenarios, involving participants from academia, technology, and public policy. == AI 2027 == In April 2025, the AI Futures Project released AI 2027, a detailed scenario forecast describing possible developments in AI between 2025 and 2027. The report was authored by Daniel Kokotajlo along with Eli Lifland, Thomas Larsen, and Romeo Dean, with editing assistance from blogger Scott Alexander. The scenario depicts very rapid progress in AI capabilities, including the development of autonomous AI systems capable of recursive self-improvement. AI 2027 presents two alternative endings: one in which international competition over advanced AI leads to catastrophic loss of human control, and another in which coordinated global action slows down development and averts imminent disaster. The authors emphasize that the narratives are hypothetical and intended as planning tools rather than literal forecasts. == Reception == AI 2027 attracted attention from technology journalists and AI researchers. Some commentators praised the report for its level of detail and its usefulness as a strategic planning exercise, while others criticized the scenario as implausibly aggressive in its timelines. The report was cited in policy discussions about AI governance. U.S. Vice President JD Vance reportedly read AI 2027 and referenced its warnings in conversations about international AI coordination. More recent reporting noted that the authors of AI 2027 had publicly revised some of their timelines. According to Kokotajlo, developments since the report's original publication suggested a slower path toward fully autonomous AI research systems than initially forecasted.
Infomax
Infomax', or the principle of maximum information preservation, is an optimization principle for artificial neural networks and other information processing systems. It prescribes that a function that maps a set of input values x {\displaystyle x} to a set of output values z ( x ) {\displaystyle z(x)} should be chosen or learned so as to maximize the average Shannon mutual information between x {\displaystyle x} and z ( x ) {\displaystyle z(x)} , subject to a set of specified constraints and/or noise processes. Infomax algorithms are learning algorithms that perform this optimization process. The principle was described by Linsker in 1988. The objective function is called the InfoMax objective. As the InfoMax objective is difficult to compute exactly, a related notion uses two models giving two outputs z 1 ( x ) , z 2 ( x ) {\displaystyle z_{1}(x),z_{2}(x)} , and maximizes the mutual information between these. This contrastive InfoMax objective is a lower bound to the InfoMax objective. Infomax, in its zero-noise limit, is related to the principle of redundancy reduction proposed for biological sensory processing by Horace Barlow in 1961, and applied quantitatively to retinal processing by Atick and Redlich. == Applications == (Becker and Hinton, 1992) showed that the contrastive InfoMax objective allows a neural network to learn to identify surfaces in random dot stereograms (in one dimension). One of the applications of infomax has been to an independent component analysis algorithm that finds independent signals by maximizing entropy. Infomax-based ICA was described by (Bell and Sejnowski, 1995), and (Nadal and Parga, 1995).
Sriram Krishnan
Sriram Krishnan (born 1984) is a tech executive and White House official, currently serving as the Senior White House Policy Advisor on Artificial Intelligence. Krishnan was named a Time Person of the Year in 2025 as an "Architect of Artificial Intelligence." He was described in Time as providing the "wake-up call that we needed" to the other AI builders, leading to "a multiyear, $500 billion initiative dubbed Stargate" to push American-made AI, as well as numerous other AI initiatives. Also in December 2025, President Trump said of Krishnan, "without him, things on AI would not function well" and cited Krishnan as the leading figure behind the American executive order on AI. As the leader of the United States' policy team regarding artificial intelligence, Krishnan plays "a significant role in shaping the administration’s approach to AI and driving measures to advance federal adoption of AI." The role calls for removing barriers to AI adoption within the government, driving vendors toward solutions suitable for federal needs, designing sensible regulation of private-sector AI, and conducting "AI diplomacy". He has stated a policy goal of "reinvigorating US dominance in emerging technologies," including AI. He also represents the United States' interests in AI abroad, such as at the Paris AI Summit. He is one of the authors of the American "AI Action Plan" released in July, 2025, which he contends is necessary to win the "existential race with China" for AI supremacy. Krishnan, a U.S. citizen born in India, is also a venture capitalist, podcaster, product manager and author. Early in his career, he led product teams at Microsoft, Twitter, Yahoo!, Facebook, and Snap. In addition to his work as an investor and technologist, he and his wife, Aarthi Ramamurthy, rose to additional prominence in 2021 as podcast hosts. He served as a general partner at the venture capital firm Andreessen Horowitz and led its London office. In 2022, Krishnan announced that he was working with Elon Musk on the rebuilding of Twitter following Musk's acquisition of the company. On December 22, 2024, US president-elect Donald Trump announced that Krishnan would be Senior White House Policy Advisor on Artificial Intelligence in his incoming administration; in 2026 he joined the National Economic Council. == Early life and education == Krishnan was born in Chennai, India. He earned his Bachelor of Technology in Information Technology from SRM University (2001–2005), moved to the United States in 2007 to join Microsoft, and became a naturalized U.S. citizen in 2016. == Career == === Early career === In 2007, he began working at Microsoft where he served as a program manager for Visual Studio. At Facebook, Krishnan built the Facebook Audience Network, a competitive platform to Google's ad technologies. At Twitter, he led product and core user experience, driving a 20% annual user growth rate and launching a redesigned home page and events experience. === Andreessen Horowitz === Krishnan was appointed a general partner of American venture capital firm Andreessen Horowitz ("a16z") in February 2021. He was anticipated to serve consumer and social markets, however he has also theorized on the impact of "deep tech" on society. In 2023 he was appointed to lead the firm's London office, its first non-US location. The office is expected to serve Web3 investments as well as AI and other fields. Krishnan announced that he would leave the firm at the end of 2024. === Social media and AI === In 2022, various news media reported that Krishnan was assisting Elon Musk in the revamp of Twitter following Musk's takeover of the company. Additional reports named Krishnan as the leading candidate for the role of CEO of the newly private company. Krishnan penned a 2023 New York Times opinion column regarding social media, AI, and related fields. He predicted a rise in the number and diversity of online spaces due to decentralization and platforms like Farcaster, Bluesky and Mastodon. === Public office === In 2024, the Financial Times reported that Krishnan was active in international affairs, reintroducing Boris Johnson to Elon Musk, following Musk's nomination to the proposed Department of Government Efficiency. Krishnan was also reported as potentially leaving a16z at the end of the year to "be jumping into something I've wanted to spend [his] energy on," which was widely reported as being related to Musk's and Vivek Ramaswamy's work at DOGE. Others reported to be involved include Joe Lonsdale, Marc Andreesen, Bill Ackman, and Travis Kalanick. On December 22, 2024, US president-elect Donald Trump announced that he would be Senior White House Policy Advisor on Artificial Intelligence in his incoming administration. On February 6, 2025, Reuters reported that Krishnan would be accompanying Vice President Vance to the Paris AI Summit, a "major artificial intelligence" event later that month. Other members of the White House Office of Science and Technology Policy would also be joining the event with around 100 other countries to "focus on AI's potential." Krishnan joined a U.S. technology policy delegation to the Middle East in advance of President Trump's visit in May 2025. Conducting "AI diplomacy," Krishnan negotiated the spread of U.S. AI technologies with Crown Prince Mohammed bin Salman of Saudi Arabia, as well as other means to strengthen bilateral trade in artificial intelligence technologies. He explained that the goal of the diplomatic mission was that "we want American A.I. to spread." Krishnan, along with David Sacks and Michael Kratsios, were credited as authors of the American AI Action Plan released in July 2025. The plan is "the administration’s most significant policy directive" regarding artificial intelligence; it calls for financing to support the global spread of American AI models and a policy to enforce neutrality in models. The Washington Post referred to the plan as a "bold action to ensure that American AI remains at the cutting edge." The AI Action Plan is a continuation of prior efforts to reduce barriers to U.S. production of AI systems and the removal of rules that were considered to hinder such growth. Later in 2025, at the POLITICO AI & Tech Summit, Krishnan called national AI development "an existential race with China." He suggested that private companies are best positioned to create new models, quipping "let them cook." He further suggested that state-by-state regulation of AI technologies may hinder national AI competitiveness. Also in 2025, at the Axios AI+ Summit, Krishnan stated that the United States and China are in a race for AI supremacy, in which the winner will be judged by market share. Winning the race is a "business strategy" to Krishnan. Krishnan was named in the 2025 Time Person of the Year article as an "AI Architect". === The Aarthi and Sriram Show and other media === In early 2021, Krishnan and his wife, Aarthi Ramamurthy, launched a Clubhouse talk show that "focuses on organic conversations on anything from startups to venture capitalism and cryptocurrencies." An early appearance by Elon Musk on the Good Time Show was described as the first show that "broke Clubhouse" by rapidly exceeding the limit of 5,000 simultaneous users. The desire to interact with a larger community led to a variety of later innovations to allow streaming and replaying of Clubhouse chats. On that episode, Elon Musk grilled Robinhood CEO Vlad Tenev regarding the GameStop trading controversy. As of December 2021, the show had over 187,000 subscribers, plus 735,000 subscribers between Krishnan and Ramamurthy's personal Clubhouse accounts. Other guests have included Facebook CEO Mark Zuckerberg, Diane von Fürstenberg, Tony Hawk, MrBeast, and A.R. Rahman. In 2022, the Good Time Show moved to YouTube. It then evolved to a podcasting format under the name The Aarthi and Sriram Show, with both audio and video content. The Hollywood Reporter reported that the podcast had received more than 1 million downloads by early 2023. == Personal life == Krishnan is married to Aarthi Ramamurthy, co-host of The Aarthi and Sriram Show (formerly the Good Time Show) and a serial entrepreneur. They met in college in 2003 through a Yahoo! chat room related to a coding project and began dating in 2006 and eloped in 2010. == Awards == Time Person of the Year - 2025
Arattai
Arattai Messenger (or simply Arattai) is an encrypted messaging service for instant messaging, voice calls, and video calls, developed by Zoho Corporation. The name Arattai means "chat" or "conversation" in Tamil. The app was soft-launched in January 2021. The app saw a sharp surge in downloads in September 2025, partially fueled by endorsements from Indian government officials. However, the app dropped from the top rankings in October 2025. == History == Arattai was initially tested internally among Zoho employees before being released publicly in early 2021. The launch coincided with a surge in interest for privacy-focused and messaging services, triggered by concerns over WhatsApp's updated terms of service. In September 2025, Arattai experienced a major surge in adoption, with daily sign-ups reportedly increasing 100-fold, from around 3,000 to more than 350,000 in three days. The surge in downloads was attributed to Zoho products being promoted by Indian government officials as part of their Make in India push for homegrown alternatives to foreign‐owned apps, amid deteriorating India–US relations. The growth temporarily strained Zoho's infrastructure, prompting rapid scaling of servers and capacity expansion. During the same period, the app reached the top position in Apple's App Store charts for the "Social Networking" category in India. The app dropped from the top ranking in late October 2025. == Reception == At launch, Arattai was positioned as a potential domestic rival to WhatsApp in India, but analysts noted that it faced challenges with encryption, ecosystem, and network effect. Critics pointed to occasional sync delays.
Semantic data model
A semantic data model (SDM) is a high-level semantics-based database description and structuring formalism (database model) for databases. This database model is designed to capture more of the meaning of an application environment than is possible with contemporary database models. An SDM specification describes a database in terms of the kinds of entities that exist in the application environment, the classifications and groupings of those entities, and the structural interconnections among them. SDM provides a collection of high-level modeling primitives to capture the semantics of an application environment. By accommodating derived information in a database structural specification, SDM allows the same information to be viewed in several ways; this makes it possible to directly accommodate the variety of needs and processing requirements typically present in database applications. The design of the present SDM is based on our experience in using a preliminary version of it. SDM is designed to enhance the effectiveness and usability of database systems. An SDM database description can serve as a formal specification and documentation tool for a database; it can provide a basis for supporting a variety of powerful user interface facilities, it can serve as a conceptual database model in the database design process; and, it can be used as the database model for a new kind of database management system. == In software engineering == A semantic data model in software engineering has various meanings: It is a conceptual data model in which semantic information is included. This means that the model describes the meaning of its instances. Such a semantic data model is an abstraction that defines how the stored symbols (the instance data) relate to the real world. It is a conceptual data model that includes the capability to express and exchange information which enables parties to interpret meaning (semantics) from the instances, without the need to know the meta-model. Such semantic models are fact-oriented (as opposed to object-oriented). Facts are typically expressed by binary relations between data elements, whereas higher order relations are expressed as collections of binary relations. Typically binary relations have the form of triples: Object-RelationType-Object. For example: the Eiffel Tower