AdTruth is a software product and the digital media division of 41st Parameter, a company headquartered in Scottsdale, Arizona, with regional offices in San Jose, California; London, England; and Munich, Germany. AdTruth allows marketers to recognize and reach target audiences across online devices. AdTruth software identifies users for targeting, tracking, performance tracking across digital media, including mobile and desktop, by analysing patterns in large numbers of advertisements served over the internet, rather than through the use of cookies. == History == AdTruth was founded in 2011 by Ori Eisen of 41st Parameter, to repurpose the company's fraud detection and prevention technology, for use within the advertising industry to accurately target intended audiences, particularly in mobile. Eisen was joined by James Lamberti in the role of vice president and general manager. In 2012 41st Parameter raised $13 million in Series D financing from Norwest Venture Partners, Kleiner Perkins Caufield & Byers, Jafco Ventures and Georgian Partners, bringing total funding to about $35 million. In May 2012, AdTruth hosted a meeting of digital media executives to discuss Apple’s UDID deprecation, with the intent of developing a device-neutral replacement standard. AdTruth joined the World Wide Web Consortium's Tracking Protection Working Group, which provides guidance for implementing and adhering to Do Not Track policies. AdTruth also worked with privacy firm Truste to create a privacy compliant Do Not Track-style mechanism for mobile. In 2013, the company Experian purchased 41st Parameter, acquiring AdTruth as part of the deal. == Product == AdTruth software helps marketers track, target and retarget consumers using more than 100 parameters, including milliseconds in differences in the internal clock setting, to recognize a particular device anonymously. AdTruth's technology uses non-UDID information to identify a wide range of devices for cookieless ad targeting. Its technology currently has about a 90 percent accuracy rate on iOS, higher on Android and desktop. AdTruth also has mobile web to app bridging capabilities as well as DeviceInsight technology, enabling marketers to identify users across mobile web and app content. 41st Parameter's patented AdTruth technology is being used by MdotM, in response to the deprecation of the UDID that included tracking and targeting capabilities. == Competitors == AdTruth's main competitor is BlueCava, which deploys a similar device-fingerprinting technology.
Resolution enhancement technology
Resolution enhancement technology (RET) is a form of image processing technology used to manipulate dot characteristics popular among laser printer and inkjet printer manufacturers. Closely related RET techniques are also used in VLSI photolithography manufacturing technology, in particular in relation to 90 nanometre technology. Resolution refers to the sharpness of image detail, smoothness of curved lines, and the faithful reproduction of an image. In both cases, RET uses pre-compensation of the image in order to try to mitigate the effects of the printing process. Among the major issues in RET in VLSI technology are the fundamental properties of a wave: amplitude, phase, and direction.
Probabilistic automaton
In mathematics and computer science, the probabilistic automaton (PA) is a generalization of the nondeterministic finite automaton; it includes the probability of a given transition into the transition function, turning it into a transition matrix. Thus, the probabilistic automaton also generalizes the concepts of a Markov chain and of a subshift of finite type. The languages recognized by probabilistic automata are called stochastic languages; these include the regular languages as a subset. The number of stochastic languages is uncountable. The concept was introduced by Michael O. Rabin in 1963; a certain special case is sometimes known as the Rabin automaton (not to be confused with the subclass of ω-automata also referred to as Rabin automata). In recent years, a variant has been formulated in terms of quantum probabilities, the quantum finite automaton. == Informal Description == For a given initial state and input character, a deterministic finite automaton (DFA) has exactly one next state, and a nondeterministic finite automaton (NFA) has a set of next states. A probabilistic automaton (PA) instead has a weighted set (or vector) of next states, where the weights must sum to 1 and therefore can be interpreted as probabilities (making it a stochastic vector). The notions states and acceptance must also be modified to reflect the introduction of these weights. The state of the machine as a given step must now also be represented by a stochastic vector of states, and a state accepted if its total probability of being in an acceptance state exceeds some cut-off. A PA is in some sense a half-way step from deterministic to non-deterministic, as it allows a set of next states but with restrictions on their weights. However, this is somewhat misleading, as the PA utilizes the notion of the real numbers to define the weights, which is absent in the definition of both DFAs and NFAs. This additional freedom enables them to decide languages that are not regular, such as the p-adic languages with irrational parameters. As such, PAs are more powerful than both DFAs and NFAs (which are famously equally powerful). == Formal Definition == The probabilistic automaton may be defined as an extension of a nondeterministic finite automaton ( Q , Σ , δ , q 0 , F ) {\displaystyle (Q,\Sigma ,\delta ,q_{0},F)} , together with two probabilities: the probability P {\displaystyle P} of a particular state transition taking place, and with the initial state q 0 {\displaystyle q_{0}} replaced by a stochastic vector giving the probability of the automaton being in a given initial state. For the ordinary non-deterministic finite automaton, one has a finite set of states Q {\displaystyle Q} a finite set of input symbols Σ {\displaystyle \Sigma } a transition function δ : Q × Σ → ℘ ( Q ) {\displaystyle \delta :Q\times \Sigma \to \wp (Q)} a set of states F {\displaystyle F} distinguished as accepting (or final) states F ⊆ Q {\displaystyle F\subseteq Q} . Here, ℘ ( Q ) {\displaystyle \wp (Q)} denotes the power set of Q {\displaystyle Q} . By use of currying, the transition function δ : Q × Σ → ℘ ( Q ) {\displaystyle \delta :Q\times \Sigma \to \wp (Q)} of a non-deterministic finite automaton can be written as a membership function δ : Q × Σ × Q → { 0 , 1 } {\displaystyle \delta :Q\times \Sigma \times Q\to \{0,1\}} so that δ ( q , a , q ′ ) = 1 {\displaystyle \delta (q,a,q^{\prime })=1} if q ′ ∈ δ ( q , a ) {\displaystyle q^{\prime }\in \delta (q,a)} and 0 {\displaystyle 0} otherwise. The curried transition function can be understood to be a matrix with matrix entries [ θ a ] q q ′ = δ ( q , a , q ′ ) {\displaystyle \left[\theta _{a}\right]_{qq^{\prime }}=\delta (q,a,q^{\prime })} The matrix θ a {\displaystyle \theta _{a}} is then a square matrix, whose entries are zero or one, indicating whether a transition q → a q ′ {\displaystyle q{\stackrel {a}{\rightarrow }}q^{\prime }} is allowed by the NFA. Such a transition matrix is always defined for a non-deterministic finite automaton. The probabilistic automaton replaces these matrices by a family of right stochastic matrices P a {\displaystyle P_{a}} , for each symbol a in the alphabet Σ {\displaystyle \Sigma } so that the probability of a transition is given by [ P a ] q q ′ {\displaystyle \left[P_{a}\right]_{qq^{\prime }}} A state change from some state to any state must occur with probability one, of course, and so one must have ∑ q ′ [ P a ] q q ′ = 1 {\displaystyle \sum _{q^{\prime }}\left[P_{a}\right]_{qq^{\prime }}=1} for all input letters a {\displaystyle a} and internal states q {\displaystyle q} . The initial state of a probabilistic automaton is given by a row vector v {\displaystyle v} , whose components are the probabilities of the individual initial states q {\displaystyle q} , that add to 1: ∑ q [ v ] q = 1 {\displaystyle \sum _{q}\left[v\right]_{q}=1} The transition matrix acts on the right, so that the state of the probabilistic automaton, after consuming the input string a b c {\displaystyle abc} , would be v P a P b P c {\displaystyle vP_{a}P_{b}P_{c}} In particular, the state of a probabilistic automaton is always a stochastic vector, since the product of any two stochastic matrices is a stochastic matrix, and the product of a stochastic vector and a stochastic matrix is again a stochastic vector. This vector is sometimes called the distribution of states, emphasizing that it is a discrete probability distribution. Formally, the definition of a probabilistic automaton does not require the mechanics of the non-deterministic automaton, which may be dispensed with. Formally, a probabilistic automaton PA is defined as the tuple ( Q , Σ , P , v , F ) {\displaystyle (Q,\Sigma ,P,v,F)} . A Rabin automaton is one for which the initial distribution v {\displaystyle v} is a coordinate vector; that is, has zero for all but one entries, and the remaining entry being one. == Stochastic languages == The set of languages recognized by probabilistic automata are called stochastic languages. They include the regular languages as a subset. Let F = Q accept ⊆ Q {\displaystyle F=Q_{\text{accept}}\subseteq Q} be the set of "accepting" or "final" states of the automaton. By abuse of notation, Q accept {\displaystyle Q_{\text{accept}}} can also be understood to be the column vector that is the membership function for Q accept {\displaystyle Q_{\text{accept}}} ; that is, it has a 1 at the places corresponding to elements in Q accept {\displaystyle Q_{\text{accept}}} , and a zero otherwise. This vector may be contracted with the internal state probability, to form a scalar. The language recognized by a specific automaton is then defined as L η = { s ∈ Σ ∗ | v P s Q accept > η } {\displaystyle L_{\eta }=\{s\in \Sigma ^{}\vert vP_{s}Q_{\text{accept}}>\eta \}} where Σ ∗ {\displaystyle \Sigma ^{}} is the set of all strings in the alphabet Σ {\displaystyle \Sigma } (so that is the Kleene star). The language depends on the value of the cut-point η {\displaystyle \eta } , normally taken to be in the range 0 ≤ η < 1 {\displaystyle 0\leq \eta <1} . A language is called η-stochastic if and only if there exists some PA that recognizes the language, for fixed η {\displaystyle \eta } . A language is called stochastic if and only if there is some 0 ≤ η < 1 {\displaystyle 0\leq \eta <1} for which L η {\displaystyle L_{\eta }} is η-stochastic. A cut-point is said to be an isolated cut-point if and only if there exists a δ > 0 {\displaystyle \delta >0} such that | v P ( s ) Q accept − η | ≥ δ {\displaystyle \vert vP(s)Q_{\text{accept}}-\eta \vert \geq \delta } for all s ∈ Σ ∗ {\displaystyle s\in \Sigma ^{}} == Properties == Every regular language is stochastic, and more strongly, every regular language is η-stochastic. A weak converse is that every 0-stochastic language is regular; however, the general converse does not hold: there are stochastic languages that are not regular. Every η-stochastic language is stochastic, for some 0 < η < 1 {\displaystyle 0<\eta <1} . Every stochastic language is representable by a Rabin automaton. If η {\displaystyle \eta } is an isolated cut-point, then L η {\displaystyle L_{\eta }} is a regular language. == p-adic languages == The p-adic languages provide an example of a stochastic language that is not regular, and also show that the number of stochastic languages is uncountable. A p-adic language is defined as the set of strings L η ( p ) = { n 1 n 2 n 3 … | 0 ≤ n k < p and 0. n 1 n 2 n 3 … > η } {\displaystyle L_{\eta }(p)=\{n_{1}n_{2}n_{3}\ldots \vert 0\leq n_{k}
DFA minimization
In automata theory (a branch of theoretical computer science), DFA minimization is the task of transforming a given deterministic finite automaton (DFA) into an equivalent DFA that has a minimum number of states. Here, two DFAs are called equivalent if they recognize the same regular language. Several different algorithms accomplishing this task are known and described in standard textbooks on automata theory. == Minimal DFA == For each regular language, there also exists a minimal automaton that accepts it, that is, a DFA with a minimum number of states and this DFA is unique (except that states can be given different names). The minimal DFA ensures minimal computational cost for tasks such as pattern matching. There are three classes of states that can be removed or merged from the original DFA without affecting the language it accepts. Unreachable states are the states that are not reachable from the initial state of the DFA, for any input string. These states can be removed. Dead states are the states from which no final state is reachable. These states can be removed unless the automaton is required to be complete. Nondistinguishable states are those that cannot be distinguished from one another for any input string. These states can be merged. DFA minimization is usually done in three steps: remove dead and unreachable states (this will accelerate the following step), merge nondistinguishable states, optionally, re-create a single dead state ("sink" state) if the resulting DFA is required to be complete. == Unreachable states == The state p {\displaystyle p} of a deterministic finite automaton M = ( Q , Σ , δ , q 0 , F ) {\displaystyle M=(Q,\Sigma ,\delta ,q_{0},F)} is unreachable if no string w {\displaystyle w} in Σ ∗ {\displaystyle \Sigma ^{}} exists for which p = δ ∗ ( q 0 , w ) {\displaystyle p=\delta ^{}(q_{0},w)} . In this definition, Q {\displaystyle Q} is the set of states, Σ {\displaystyle \Sigma } is the set of input symbols, δ {\displaystyle \delta } is the transition function (mapping a state and an input symbol to a set of states), δ ∗ {\displaystyle \delta ^{}} is its extension to strings (also known as extended transition function), q 0 {\displaystyle q_{0}} is the initial state, and F {\displaystyle F} is the set of accepting (also known as final) states. Reachable states can be obtained with the following algorithm: Assuming an efficient implementation of the state sets (e.g. new_states) and operations on them (such as adding a state or checking whether it is present), this algorithm can be implemented with time complexity O ( n + m ) {\displaystyle O(n+m)} , where n {\displaystyle n} is the number of states and m {\displaystyle m} is the number of transitions of the input automaton. Unreachable states can be removed from the DFA without affecting the language that it accepts. == Nondistinguishable states == The following algorithms present various approaches to merging nondistinguishable states. === Hopcroft's algorithm === One algorithm for merging the nondistinguishable states of a DFA, due to Hopcroft (1971), is based on partition refinement, partitioning the DFA states into groups by their behavior. These groups represent equivalence classes of the Nerode congruence, whereby every two states are equivalent if they have the same behavior for every input sequence. That is, for every two states p1 and p2 that belong to the same block of the partition P, and every input word w, the transitions determined by w should always take states p1 and p2 to either states that both accept or states that both reject. It should not be possible for w to take p1 to an accepting state and p2 to a rejecting state or vice versa. The following pseudocode describes the form of the algorithm as given by Xu. Alternative forms have also been presented. The algorithm starts with a partition that is too coarse: every pair of states that are equivalent according to the Nerode congruence belong to the same set in the partition, but pairs that are inequivalent might also belong to the same set. It gradually refines the partition into a larger number of smaller sets, at each step splitting sets of states into pairs of subsets that are necessarily inequivalent. The initial partition is a separation of the states into two subsets of states that clearly do not have the same behavior as each other: the accepting states and the rejecting states. The algorithm then repeatedly chooses a set A from the current partition and an input symbol c, and splits each of the sets of the partition into two (possibly empty) subsets: the subset of states that lead to A on input symbol c, and the subset of states that do not lead to A. Since A is already known to have different behavior than the other sets of the partition, the subsets that lead to A also have different behavior than the subsets that do not lead to A. When no more splits of this type can be found, the algorithm terminates. Lemma. Given a fixed character c and an equivalence class Y that splits into equivalence classes B and C, only one of B or C is necessary to refine the whole partition. Example: Suppose we have an equivalence class Y that splits into equivalence classes B and C. Suppose we also have classes D, E, and F; D and E have states with transitions into B on character c, while F has transitions into C on character c. By the Lemma, we can choose either B or C as the distinguisher, let's say B. Then the states of D and E are split by their transitions into B. But F, which doesn't point into B, simply doesn't split during the current iteration of the algorithm; it will be refined by other distinguisher(s). Observation. All of B or C is necessary to split referring classes like D, E, and F correctly—subsets won't do. The purpose of the outermost if statement (if Y is in W) is to patch up W, the set of distinguishers. We see in the previous statement in the algorithm that Y has just been split. If Y is in W, it has just become obsolete as a means to split classes in future iterations. So Y must be replaced by both splits because of the Observation above. If Y is not in W, however, only one of the two splits, not both, needs to be added to W because of the Lemma above. Choosing the smaller of the two splits guarantees that the new addition to W is no more than half the size of Y; this is the core of the Hopcroft algorithm: how it gets its speed, as explained in the next paragraph. The worst case running time of this algorithm is O(ns log n), where n is the number of states and s is the size of the alphabet. This bound follows from the fact that, for each of the ns transitions of the automaton, the sets drawn from Q that contain the target state of the transition have sizes that decrease relative to each other by a factor of two or more, so each transition participates in O(log n) of the splitting steps in the algorithm. The partition refinement data structure allows each splitting step to be performed in time proportional to the number of transitions that participate in it. This remains the most efficient algorithm known for solving the problem, and for certain distributions of inputs its average-case complexity is even better, O(n log log n). Once Hopcroft's algorithm has been used to group the states of the input DFA into equivalence classes, the minimum DFA can be constructed by forming one state for each equivalence class. If S is a set of states in P, s is a state in S, and c is an input character, then the transition in the minimum DFA from the state for S, on input c, goes to the set containing the state that the input automaton would go to from state s on input c. The initial state of the minimum DFA is the one containing the initial state of the input DFA, and the accepting states of the minimum DFA are the ones whose members are accepting states of the input DFA. === Moore's algorithm === Moore's algorithm for DFA minimization is due to Edward F. Moore (1956). Like Hopcroft's algorithm, it maintains a partition that starts off separating the accepting from the rejecting states, and repeatedly refines the partition until no more refinements can be made. At each step, it replaces the current partition with the coarsest common refinement of s + 1 partitions, one of which is the current one and the rest of which are the preimages of the current partition under the transition functions for each of the input symbols. The algorithm terminates when this replacement does not change the current partition. Its worst-case time complexity is O(n2s): each step of the algorithm may be performed in time O(ns) using a variant of radix sort to reorder the states so that states in the same set of the new partition are consecutive in the ordering, and there are at most n steps since each one but the last increases the number of sets in the partition. The instances of the DFA minimization problem that cause the worst-case behavior are the same as for Hopcroft's algorithm. The number of steps th
The Best Free AI Subtitle Generator for Beginners
In search of the best AI subtitle generator? An AI subtitle generator is software that uses machine learning to help you get more done — it turns a rough idea into a polished result in seconds. When choosing one, weigh output quality, pricing, export formats, and how well it fits the tools you already use. Whether you are a beginner or a pro, the right AI subtitle generator slots into your workflow and pays for itself fast. Below we compare features, pricing, and real output so you can choose with confidence.
Automated essay scoring
Automated essay scoring (AES) is the use of specialized computer programs to assign grades to essays written in an educational setting. It is a form of educational assessment and an application of natural language processing. Its objective is to classify a large set of textual entities into a small number of discrete categories, corresponding to the possible grades, for example, the numbers 1 to 6. Therefore, it can be considered a problem of statistical classification. Several factors have contributed to a growing interest in AES. Among them are cost, accountability, standards, and technology. Rising education costs have led to pressure to hold the educational system accountable for results by imposing standards. The advance of information technology promises to measure educational achievement at reduced cost. The use of AES for high-stakes testing in education has generated significant backlash, with opponents pointing to research that computers cannot yet grade writing accurately and arguing that their use for such purposes promotes teaching writing in reductive ways (i.e. teaching to the test). == History == Most historical summaries of AES trace the origins of the field to the work of Ellis Batten Page. In 1966, he argued for the possibility of scoring essays by computer, and in 1968 he published his successful work with a program called Project Essay Grade (PEG). Using the technology of that time, computerized essay scoring would not have been cost-effective, so Page abated his efforts for about two decades. Eventually, Page sold PEG to Measurement Incorporated. By 1990, desktop computers had become so powerful and so widespread that AES was a practical possibility. As early as 1982, a UNIX program called Writer's Workbench was able to offer punctuation, spelling and grammar advice. In collaboration with several companies (notably Educational Testing Service), Page updated PEG and ran some successful trials in the early 1990s. Peter Foltz and Thomas Landauer developed a system using a scoring engine called the Intelligent Essay Assessor (IEA). IEA was first used to score essays in 1997 for their undergraduate courses. It is now a product from Pearson Educational Technologies and used for scoring within a number of commercial products and state and national exams. IntelliMetric is Vantage Learning's AES engine. Its development began in 1996. It was first used commercially to score essays in 1998. Educational Testing Service offers "e-rater", an automated essay scoring program. It was first used commercially in February 1999. Jill Burstein was the team leader in its development. ETS's Criterion Online Writing Evaluation Service uses the e-rater engine to provide both scores and targeted feedback. Lawrence Rudner has done some work with Bayesian scoring, and developed a system called BETSY (Bayesian Essay Test Scoring sYstem). Some of his results have been published in print or online, but no commercial system incorporates BETSY as yet. Under the leadership of Howard Mitzel and Sue Lottridge, Pacific Metrics developed a constructed response automated scoring engine, CRASE. Currently utilized by several state departments of education and in a U.S. Department of Education-funded Enhanced Assessment Grant, Pacific Metrics’ technology has been used in large-scale formative and summative assessment environments since 2007. Measurement Inc. acquired the rights to PEG in 2002 and has continued to develop it. In 2012, the Hewlett Foundation sponsored a competition on Kaggle called the Automated Student Assessment Prize (ASAP). 201 challenge participants attempted to predict, using AES, the scores that human raters would give to thousands of essays written to eight different prompts. The intent was to demonstrate that AES can be as reliable as human raters, or more so. The competition also hosted a separate demonstration among nine AES vendors on a subset of the ASAP data. Although the investigators reported that the automated essay scoring was as reliable as human scoring, this claim was not substantiated by any statistical tests because some of the vendors required that no such tests be performed as a precondition for their participation. Moreover, the claim that the Hewlett Study demonstrated that AES can be as reliable as human raters has since been strongly contested, including by Randy E. Bennett, the Norman O. Frederiksen Chair in Assessment Innovation at the Educational Testing Service. Some of the major criticisms of the study have been that five of the eight datasets consisted of paragraphs rather than essays, four of the eight data sets were graded by human readers for content only rather than for writing ability, and that rather than measuring human readers and the AES machines against the "true score", the average of the two readers' scores, the study employed an artificial construct, the "resolved score", which in four datasets consisted of the higher of the two human scores if there was a disagreement. This last practice, in particular, gave the machines an unfair advantage by allowing them to round up for these datasets. In 1966, Page hypothesized that, in the future, the computer-based judge will be better correlated with each human judge than the other human judges are. Despite criticizing the applicability of this approach to essay marking in general, this hypothesis was supported for marking free text answers to short questions, such as those typical of the British GCSE system. Results of supervised learning demonstrate that the automatic systems perform well when marking by different human teachers is in good agreement. Unsupervised clustering of answers showed that excellent papers and weak papers formed well-defined clusters, and the automated marking rule for these clusters worked well, whereas marks given by human teachers for the third cluster ('mixed') can be controversial, and the reliability of any assessment of works from the 'mixed' cluster can often be questioned (both human and computer-based). == Different dimensions of essay quality == According to a recent survey, modern AES systems try to score different dimensions of an essay's quality in order to provide feedback to users. These dimensions include the following items: Grammaticality: following grammar rules Usage: using of prepositions, word usage Mechanics: following rules for spelling, punctuation, capitalization Style: word choice, sentence structure variety Relevance: how relevant of the content to the prompt Organization: how well the essay is structured Development: development of ideas with examples Cohesion: appropriate use of transition phrases Coherence: appropriate transitions between ideas Thesis Clarity: clarity of the thesis Persuasiveness: convincingness of the major argument == Procedure == From the beginning, the basic procedure for AES has been to start with a training set of essays that have been carefully hand-scored. The program evaluates surface features of the text of each essay, such as the total number of words, the number of subordinate clauses, or the ratio of uppercase to lowercase letters—quantities that can be measured without any human insight. It then constructs a mathematical model that relates these quantities to the scores that the essays received. The same model is then applied to calculate scores of new essays. Recently, one such mathematical model was created by Isaac Persing and Vincent Ng. which not only evaluates essays on the above features, but also on their argument strength. It evaluates various features of the essay, such as the agreement level of the author and reasons for the same, adherence to the prompt's topic, locations of argument components (major claim, claim, premise), errors in the arguments, cohesion in the arguments among various other features. In contrast to the other models mentioned above, this model is closer in duplicating human insight while grading essays. Due to the growing popularity of deep neural networks, deep learning approaches have been adopted for automated essay scoring, generally obtaining superior results, often surpassing inter-human agreement levels. The various AES programs differ in what specific surface features they measure, how many essays are required in the training set, and most significantly in the mathematical modeling technique. Early attempts used linear regression. Modern systems may use linear regression or other machine learning techniques often in combination with other statistical techniques such as latent semantic analysis and Bayesian inference. The automated essay scoring task has also been studied in the cross-domain setting using machine learning models, where the models are trained on essays written for one prompt (topic) and tested on essays written for another prompt. Successful approaches in the cross-domain scenario are based on deep neural networks or models that combine deep and shallow features. == Criteria for success == Any method of a
Forrest N. Iandola
Forrest N. Iandola is an American computer scientist specializing in efficient AI. == Career == Iandola earned a PhD in Electrical Engineering and Computer Science from UC Berkeley in 2016, advised by Kurt Keutzer. As part of his dissertation, he co-authored SqueezeNet, a deep neural network for image classification optimized for smartphones and other mobile devices. Iandola and Keutzer went on to co-found DeepScale. The firm squeezes deep neural networks onto low-cost automotive-grade processors for use in driver assistance systems. Tesla acquired DeepScale in 2019. In 2020, he co-authored SqueezeBERT, an efficient neural network for natural language processing. In 2022, he joined Meta as an AI research scientist. His research at Meta includes developing efficient AI models, such as EfficientSAM and MobileLLM.