User:Penrose Delta/sandbox

From Wikipedia, the free encyclopedia


bold

link

same link with a different name

indent

thing to be cited[1]

things typed with an enter here > <that are on the same line

Pythia (model suite)[edit]

Pythia (/ˈpɪθiə/), named after the Oracle of Delphi, is a suite of decoder-only, autoregressive large language models (LLMs) all trained on public data seen in the exact same order and ranging in size from 70 million to 12 billion parameters.[2] It was officially released by EleutherAI on April 3, 2023 with the goal of "promot[ing] scientific research on large language models".[3] The Pythia model suite contains 8 base model sizes, each trained on 2 different versions of the Pile, creating 16 total models with 154 intermediate training checkpoints available to examine for each.[4]

Creation[edit]

LLMs store information in a way that is exceedingly difficult to understand,[5] leading to the need for interpretability. Pythia was created to facilitate research in many areas of this field by creating the first suite of models to satisfy three key properties.[3]

  • The models span several orders of magnitude of model scale.
  • All of the models were trained on the same data in the same order.
  • The data and intermediate checkpoints are publicly available for study.

Architecture[edit]

The model architecture and hyperparameters largely follow GPT-3, with a few notable deviations based on recent advances in best practices for large scale language modeling.[2] The hyperparameters for each of the 8 starting models are displayed in the table below.[3]

Model hyperparameters
Parameters Transformer layers Attention heads Comparable models
70,000,000 6 8
160,000,000 12 12 GPT-Neo 125M, OPT-125M
410,000,000 24 16 OPT-350M
1,000,000,000 16 8
1,400,000,000 24 16 GPT-Neo 1.3B, OPT-1.3B
2,800,000,000 32 32 GPT-Neo 2.7B, OPT-2.7B
6,900,000,000 32 32 OPT-6.7B
12,000,000,000 36 40

The two datasets used in training were the Pile and a version of the Pile that had undergone data deduplication.[3] Each of the 16 resulting models was recorded during training at 154 publicly available checkpoints at initialization (0 steps of training), then at step 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1000, 2000, 3000, and so on at every 1000 steps.[4] The space between two 1000-step checkpoints corresponds to just under 1 epoch on the original Pile and about 1.5 epochs on the deduplicated Pile (which contains 207 billion tokens in 1 epoch).[3]

Applications[edit]

The Pythia model suite has been used to demonstrate that modifying the majority of pronouns in the last 7% and 21% of model training data from masculine to feminine can be used to successfully reduce gender bias.[3]

The model suite has also been used to study the process of memorization, in which an LLM recites back a string from its training data. Training data may sometimes include personally identifiable information, so learning how to predict which strings might be memorized (and thus how to prevent undesirable strings from being memorized) is a matter of safety and privacy.[6] It has been used to show that the location of a string within a dataset does not influence its probability of being memorized[3] and that

[7]

While previous research had shown that LLMs are able to more accurately apply knowledge from their training data when those terms are more prevalent within the training data, that research had been done on the fully trained GPT-J and GPT-3 and could not evaluate how that behavior emerged throughout training.[8] Pythia's intermediate checkpoints were able to show that models don't become able to apply knowledge well until 65 thousand training steps in.[3] The variety of model sizes was able to show that only larger models (2.8 billion parameters and up) were able to learn associated information at all.[3]

Criticism (of the Common Crawl)[edit]

The Common Crawl relies on an opt-out model of consent that relies on websites knowing that it exists and how to communicate a desire to opt out.[9][10] In many cases, work has been taken without the knowledge or consent of the content creators or site hosts.[11][12]

Once content is taken, the Common Crawl will only remove it if it infringes someone's copyright or intellectual property and only if that person submits a request for removal that complies with Title 17, United States Code, Section 512(c)(3).[10] However, the Common Crawl corpus contains petabytes of data,[13] the smallest pieces of which are hundreds of terabytes,[9] all of which is concatenated into a format that is unstructured and designed to be inaccessible to humans[9] and is far more than most computers can handle.[14] US copyright law places the burden on the complainant to not only find but give instructions for finding the infringing material,[15] so submitting such a claim is difficult or impossible for a dataset of this nature for the vast majority of content creators.[9][16] By hosting their content this way, the Common Crawl takes advantage of what is essentially a legal loophole.[17]

  • Terms of Use do not permit sub-licensing. Is this required for use in AI?
  • Terms of Use note that their content may be subject to separate terms of use or terms of service from the owners and require that you follow any technological implementations restricting how you may used that content, including the use of robot.txt files and NOFOLLOW meta-tags. I don't think most AIs check any of this.

Special Education in Japan[edit]

In Japan, special needs education (Japanese: 特別支援教育, tokubetsu shien kyōiku, lit. special support education) is defined by the Japanese Ministry of Education, Culture, Sports, Science and Technology (MEXT) (Japanese: 文部科学省, Monbu Kagaku Shō, lit. Ministry of Letters and Science) as "education for students with disabilities, in consideration of their individual educational needs, which aims at full development of their capabilities and at their independence and social participation."[18] In keeping with the general principle of equity and inclusion in education, the school system attempts to integrate students into the regular school system to the greatest extent possible.[18]

Current Procedure[edit]

Location[edit]

In terms of learning environment, there are four main arrangements these plans can take:[19]

  • accommodations within regular classrooms,
  • resource rooms (Japanese: 通級指導学級, tsūkyū shidō gakkyū, lit. general guidance class),[20] which provide part-time instruction in certain life skills within the same school,
  • special classes (Japanese: 中よし or 中よし学究, nakayoshi or nakayoshi gakkyū, lit. close friends or close friends class),[21][22][23] which are separate from regular classrooms but located within the same school, and
  • special schools (Japanese: 特別支援学校, tokubetsu shien gakkō, lit. special support school), which are separate from regular schools entirely.

In addition, an Individualized Instruction Plan (Japanese: 個別の指導計画, Kobetsu no Shidō Keikaku, lit. Individualized Guidance Plan) can be developed for any student with special needs.[24] However, IIPs are only mandated at special needs schools, so many teachers at regular schools have little to no experience reviewing or developing IIPs.[18] Individualized Instruction Plans (IIPs) were modeled off of the American system of Individualized Education Programs (IEPs) and are essentially equivalent to them.[25]

Where possible, students with special needs are fully included in the mainstream classroom, with accommodations or modifications as needed.[19] Additionally, if a parent requests that their student be placed in a regular classroom, that request is generally honored.[22]

Resource rooms are used part-time for specialized instruction. These students spend the most of the day in regular classrooms but occasionally have private lessons or are taught in small groups in resource rooms. They may take anywhere from one to several classes there per week.[20][26] Resource rooms are geared toward helping the student learn non-academic skills called "independence activities" (Japanese: 自立活動, jiritsu katsudō, lit. independence activities), such as social skills or how to study in a way compatible with their disability.[20]

For students who require more intensive help, special classes are available. They run in the same schools as regular classes, but they are not assigned a grade level.[21] Special classes catering specifically to children with health problems in addition to other disabilities are also sometimes set up as branch classes in children's hospitals.[26] In Japan, education is organized and standardized at national level. The MEXT publishes curriculum guidelines (Japanese: 学習指導要領, gakushū shidō yōryō, lit. courses for study guidelines) that specify the materials to be taught at all elementary, junior, and senior high schools in Japan, whether public or private, and which is legally enforceable.[27] Teaching of special education classes is exempted from these requirements,[24] and there are actually separate curriculum guidelines for special needs education.[28] While students in special classes take the same assessments as students in regular classrooms where possible, teachers of special classes have the discretion to deviate from the normal curriculum guidelines if necessary in accordance with a child’s individual needs and the special needs education curriculum guidelines.[19]

Special schools are reserved for students with severe disabilities who cannot be accommodated in their local school.[19] For example, major hospitals often have schools for children with health problems, and teachers may even visit a student's home to provide home tutoring.[24] Special schools provide education according to phase (kindergarten, elementary, lower secondary, and upper secondary education) but do not always use the same grading or marking systems as regular schools. Where possible, assessment for pupils in special schools is the same as that in regular schools, but similar to special classes, they are not bound to the normal curriculum guidelines.[19] Students in regular schools are typically assessed with a score of 1 to 3 in elementary schools, 1 to 5 in lower secondary school, and 1 to 5 in each subject in high school, where 1 is the lowest score and indicates a failing grade. They are assessed according to four metrics (Japanese: 観点, kanten, lit. viewpoints):[19]

  • interest, motivation, and attitude,
  • thinking and judgement,
  • skills and expression, and
  • knowledge and understanding.

Special schools do not use the same grade system, and students are instead assessed individually in accordance with their IIPs.[19]

Students from regular classes are encouraged to take advantage of opportunities for contact with disabled students. Special schools are also encouraged to participate in exchange (Japanese: 交流, kōryū, lit. cultural exchange or interaction) programs with their peers.[24]

Content[edit]

As a general rule globally, in regular education, the curriculum focuses on academic subjects with the assumption that life skills such as social skills, time management, emotional control, and how to learn, will be acquired naturally through the course of the practice of these skills necessary to learn academic subjects and function in a school setting. In Japanese special needs education for those with intellectual and developmental disabilities, this assumption is reversed. These curriculum guidelines divide elementary school (1st through 6th grade) into three developmental levels, junior high school (7th through 9th) into two, and high school (10th through 12th) into two.

  • Elementary school level 1: Children experience, notice and pay attention, gain interest, and steadily acquire basic actions, with direct help from teachers.
  • Elementary school level 2: Children imitate teachers’ actions and movements, play, take action with a purpose, and acquire basic socialized behavior, with direct help from teachers through the use of language.
  • Elementary school level 3: Children notice their situations and their patterns, proactively engage in activities, and acquire behavior suitable for social life.
  • Junior high school level 1: Students engage in activities in a proactive manner, use what they have experienced, consider the order of things, and learn the basics of daily life and social life.
  • Junior high school level 2: Students acquire the basics of daily life, social life, and occupational life for the future. Students engage in activities in a proactive manner, make choices in accordance with a purpose, creatively deal with various situations, and acquire skills with a view to future work life.
  • High school level 1: This stage mainly involves students’ family life, social life, and work life after graduation according to their age and based on the contents of level 2 of the junior high school division, as well as their life experience so far. Students will learn proactively, acquire basic life habits, social skills, and professional abilities with a view to life after graduation.
  • High school level 2: This stage, based on level 1 of the high school division, targets students whose disabilities are relatively minor. It offers practical and developed contents that consider family life after graduation, social life and work life. Students will learn proactively, acquire basic life habits necessary for life after graduation, social skills, and professional abilities.

Thus, in summary, Japanese special needs education provides a multi-track curriculum that consists of regular subjects, subjects for children with intellectual and developmental disabilities, and independence activities that are mixed and matched as needed.

Beyond High School[edit]

In order to prepare for their students' futures, many special high schools offer transition programs to assist their students in seeking jobs.[24] Similar to the American Individual Transition Plans (ITPs), which lay out a plan to assist the student in transitioning from school life to work, many Japanese schools have Individual Transition Support Plans (Japanese: 個別移行支援計画, Kobetsu Ikō Shien Keikaku, lit. Individual Transition Support Plan).[25] Transition support is split into two phases: before and after graduation.

Before graduation, Japanese special high schools provide transition support in the form of career guidance (Japanese: 進路指導, shinro shidō, lit. course guidance). This typically includes career education, job training, and career counseling meetings. Job training typically takes the form of unpaid internships, work-site visits, and general study units. Career counseling meetings involve a student meeting with both teachers and guardians to discuss their options for the future. The biggest problems at this phase tend to arise from disagreements between students, teachers, and parents regarding what constitutes realistic goals.[25]

After graduation, some Japanese special high schools perform check-ins with their students. However, this is not formally mandated, so many schools do not provide the services at all, and those that do suffer from inconsistent budgeting and difficulties during succession when, for instance, a teacher retires.[25]

There are also accommodations given on university entrance exams. The National Center Test for University Admissions (Japanese: 大学入試センター試験, Daigaku Nyūshi Sentā Shiken, lit. University Entrance Exam Center Test) was a standardized test used as an admissions exam by many universities in Japan. While it was replaced by the Common Test for University Admissions (Japanese: 大学入学共通テスト, Daigaku Nyūgaku Kyōtsū Tesuto, lit. Common Test for University Admissions) in 2021, it offered a number of accommodation options for those who needed them, including the provision of examination papers in Braille or an enlarged print, the provision of a reading glass, consideration of the candidate’s seating position in the examination hall, the provision of additional time to answer questions, the assistance of a sign language interpreter, provision of written guidance and/or hearing aids, exemption from an aural/oral test, the provision of an assistant reader or writer for the candidate, and/or the provision of a separate room in which to take the examination.[19] Similar arrangements are provided for normal end-of-term examinations and other standardized tests,[19] so the Common Test for University Admissions might offer the same.

Abuse[edit]

Across the world, students with special needs tend to suffer from a higher rate of abuse than their peers.[29] Japanese special education schools are no exception to this trend. One measure employed at such schools to prevent abuse is educational programs for staff.[30]

National Institute of Special Needs Education (NISE)[edit]

The Japanese National Institute of Special Needs Education (NISE) (Japanese: 国立特別支援教育総合研究所, Kokuritsu Tokubetsu Shien Kyōiku Sōgōkenkyūjo, lit. National Special Support Education Research Institute) is an organization that conducts research, provides training and support, and disseminates information regarding the education of children with disabilities.[24] It aims to improve the education of children with disabilities in Japan and develop an inclusive education system, in which children with and without disabilities study together, and which allows each child to fully demonstrate their abilities.[31]

It began when it was established as an affiliated institution of MEXT in 1971.[24][31] At the time, it was named the National Institute of Special Education (Japanese: 国立特殊教育総合研究所, Kokuritsu Tokushu Kyōiku Sōgōkenkyūjo, lit. National Special Education Research Institute). In 2001, NISE was reestablished as an independent administrative agency under MEXT's jurisdiction. The acronym is a holdover from its original name, and it did not take its current name until April 1, 2007. The name change was partially motivated by revisions made to the School Education Act in 2006 as part of an overall shift in focus from providing accommodations according to disability category to providing accommodations according to the needs of the individual.[31]

The institute is located at 5-1-1 Nobi, in Yokosuka, Kanagawa Prefecture.[24] The current (as of 2023) President is Kazushige Shishido (Japanese: 宍戸和成 Shishido Kazushige).

History[edit]

The Japanese system began with special schools designed according to the disabilities of the students. Six categories were commonly used: visual impairment, hearing impairment, intellectual disabilities, physical disabilities, health impairments, and multiple disabilities. The paradigm shifted to special needs classrooms within regular schools and, later, to ensuring access to regular schools for children with disabilities. The overall trend has thus been from separation to inclusion, and more recently, the focus of the inclusive education system has shifted from the location in which classes and support are offered, to the content that best meets the needs of students with disabilities of every type.[28]

In 1878, Tashirō Furukawa (Japanese: 古河太四郎, Furukawa Tashirō) founded the Kyōto School for the Deaf and Blind (Japanese: Kyōto Moain), which was later divided into Kyōto Prefectural School for the Blind and Kyōto Prefectural School for the Deaf.[32][33] Between 1879 and 1881 Ōsaka, Tōkyō, and Kanazawa founded schools for the blind and deaf as well.[32]

In 1891, Japan officially separated its schools by grade level.[34]

In 1947, Japan passed the School Education Act, which overhauled the entire education system. It required the provision of free, compulsory education for six years of elementary school and three years of junior high school.[28]

In 1979, compulsory education was mandated for all children with disabilities, and, as of 2002, only 0.001% of children are allowed postponement of or exemption from school education because of their disabilities.[24] However, this law mandated that children enter a special school education system for children with disabilities, when such children had previously been allowed to attend regular classes. Paradoxically, the new mandate forcibly transferred some children with disabilities to schools exclusively for children with disabilities. It included a list of thresholds for degree of disability with no exceptions allowed beyond that point.[28] Children below these thresholds were allowed to remain in the regular school system in special classes.

On January 28, 1993, MEXT issued another revision systematizing the use of resource rooms for students with slight disabilities at the beginning of the next academic year (April 1, 1993). The special needs support in resource rooms was deemed to be suitable for educating students with speech and language disorders, emotional disturbances, low vision, hearing problems, and students with other difficulties.[28]

In June of 1994, representatives of 92 governments and 25 international organizations formed the World Conference on Special Needs Education, held in Salamanca, Spain. They released the UNESCO Salamanca Statement regarding the education of all disabled children, which called for inclusion to be the norm. [35]

In 2002, the Ordinance for the Enforcement of the School Education Act was revised to give flexibility to local governments in deciding where to place students with disabilities.[28]

On March 31, 2006, MEXT released another revision effective the following day. This ordinance included students with learning disorders and ADHD as examples of the types of students with disabilities for whom special needs support in resource rooms was considered appropriate. It also moved autism from the category of emotional disturbances to its own category so that the numbers could be counted separately.[28]

On December 13, 2006, the United Nations General Assembly drafted the Convention on the Rights of Persons with Disabilities, an international human rights treaty intended to protect the rights and dignity of persons with disabilities.[36] Article 24 requires the free, compulsory, inclusive education of children with disabilities through secondary education.[36] Japan signed the treaty in 2007,[37] resulting in an official shift in focus from from providing accommodations according to disability category to providing accommodations according to the needs of the individual.[28] The School Education Law was amended that same year, which cause several major changes: the addition of a special needs education coordinator in all schools, the addition of a support assistant for special needs education in regular schools, the additional role for special schools to provide advice and supports to students with disabilities at local regular schools, the hearing of parents and guardians' opinions by the local education board regarding school placement, and the requirement of a comprehensive teacher’s special needs education certification system.[37]

On May 3, 2008, the UN brought the Convention on the Rights of Persons with Disabilities into effect.[38] In order to meet the requirements of Article 24, the Ordinance for Enforcement of the School Education Act was revised again in August 2013. The new framework of school selection specified that schools for children with disabilities should be selected in comprehensive consideration of a number of criteria:

  • the disabilities in question,
  • the educational needs of the child,
  • the wishes of the child,
  • the wishes of their parents or guardians,
  • the opinions of experts in education, medicine, and psychology, and
  • the conditions and resources of the specific schools and communities available.

Later, opportunities for hearing the opinions of experts and parents or guardians were expanded. Eventually, it was decided that schools would be selected by education committees after respecting the opinions of the children with disabilities and their parents or guardians. Regular school became the default placement, and the previous list of thresholds instead became part of the conditions that children must meet when applying for special needs schools.[28] These efforts were successful, and in 2014, Japan officially ratified the Convention on the Rights of Persons with Disabilities.[37]

In 2017, MEXT released additional guidelines for creating an individualized curriculum for a child attending a regular school who needed more help than the normal curriculum guidelines could provide but less than what was typically expected with the special needs education curriculum. These guidelines provide instructions on how to properly create a mix of the two that meets the child's needs. In particular, content can be replaced by the content taught at a lower grade if necessary, content can be replaced by the content created for students with intellectual and developmental disabilities if necessary, subjects can be replaced by independence activities if necessary, and students who have completed the content of the special needs education curriculum guidelines for those with intellectual and developmental disabilities for their school level are free (and encouraged) to learn as much from the regular curriculum guidelines as they can.[28]

See also[edit]

Creation of the KLOE (experiment) page[edit]

This page is about the particle detector and experiment. For the AM radio station in Kansas, see KLOE.

KLOE (or the
K0
L LOng Experiment) was both an experiment studying ϕ meson decays, and the particle detector used to conduct it. It was located in the DAϕNE collider at the INFN Frascati National Laboratory in Frascati, Italy. It ceased operation in 2006 and was replaced by the KLOE-2 detector, which began operation in 2014, and continues to operate to this day.[39] Makoto Kobayashi and Toshihide Maskawa used data from the original KLOE experiment to win the Nobel Prize in Physics in 2008.[40]

Etymology[edit]

Both the DAϕNE collider and the KLOE detector were named after the two titular characters of the ancient Greek play Daphnis and Chloe, written in the second century AD.[41] In the story, the two grow up and fall in love, experiencing various hardships before living happily ever after. The DAϕNE collider was designed with the KLOE experiment as its primary goal, leading to the two to be named as a pair.

KLOE and the Nobel Prize[edit]

The KLOE experiment was the first experiment performed by the DAϕNE collider,[39] and it began in ernest when the detector began taking data in 2000 and ended when data collection stopped in 2006.[42] [39] The KLOE detector was cylindrical in shape. It had a length of 6 meters and a diameter of 7 meters and was composed of a drift chamber surrounded by a concentric electromagnetic calorimeter, both of which were kept within a constant magnetic field.[43] The interior drift chamber had a length of 3.3 meters and a diameter of 4 meters, within which it contained 52,000 wires, making it the largest drift chamber ever constructed at the time.[43] The computer interpreting its data was able to calculate reconstructed particle trajectories with a precision within 0.3%.[43] The electromagnetic calorimeter had a 4.5 meters and a diameter of 4 meters. It used alternating layers of lead with 15,000 kilometers of scintillating fibers before passing the energy from the fibers through 4880 photomultipliers. It was able to determine the energy released by a given particle to within 15% precision, and was able to distinguish between particles occurring at least 0.2 nanoseconds apart, but was limited to the computer's ability to calculate a maximum of 2000 events per second.[43] The detector was designed to witness the decays of
K0
L mesons that were created by colliding electrons and positrons at high speeds to generate large numbers of
ϕ
 mesons, 34.2%±0.4% of which then decay into the
K0
S
K0
L pair, following the second most common decay mode.[44]

Data from the original experiment allowed Makoto Kobayashi and Toshihide Maskawa to give the most precise measurement of one of the elements of the quark mixing matrix currently available. They then used this value to discover "the origin of of the broken symmetry which predicts the existence of at least three families of quarks in nature", which won them the Nobel Prize in Physics in 2008.[40]

KLOE-2[edit]

KLOE-2 began taking data in November of 2014 and is scheduled to continue taking data until at least 2018. Its first run, Run-I was began in November of 2014 and continued until July of 2015, observing a total of 1 billion neutral kaon decays.[39] The second experiment, Run-II is still in progress and aims to reach 5 billion such observations. Its drift chamber has the same dimensions as KLOE. It also uses lead and scintillating fibers and the same number of photomultiplier tubes. It uses a magnetic field strength of 0.52T.[41]

My edits to the Phi meson page[edit]

phi meson
Feynman diagram of the most common ϕ meson decay
Composition
ϕ0
:
s

s
StatisticsBosonic
FamilyMeson
InteractionsStrong, Weak
Symbol
ϕ
,
ϕ0
AntiparticleSelf
Mass1019.445±0.020 MeV/c2
Electric charge0

In particle physics, the phi meson or ϕ meson is a vector meson formed of a strange quark and a strange antiquark. This gives it the distinction of being the only known particle that has a strangeness of 0 despite containing strange flavored quarks. It was the ϕ meson's unusual propensity to decay into
K0
 and
K0
 that led to the discovery of the OZI rule. It has a mass of 1019.445±0.020 MeV/c2 and a mean lifetime of 1.55±0.01 × 10−22s.

Properties[edit]

Particle name Particle
symbol 		Antiparticle
symbol 		Quark
content 		Rest mass (MeV/c2) IG JPC S C B' Mean lifetime (s) Commonly decays to

(>5% of decays)

Phi meson[44]
ϕ
(1020) 		Self
s

s
 		1,019.445 ± 0.020 0− 1−− 0 0 0 1.55 ± 0.01 × 10−22[f]
K+
 +
K−
 or

K0
S +
K0
L or
(
ρ
 +
π
) / (
π+
 +
π0
 +
π−
)

The most common decay paths for the ϕ meson are
K+

K−
 at 48.9%±0.5%,
K0
S
K0
L at 34.2%±0.4%, and various indistinguishable combinations of
ρ
s and pions at 15.3%±0.3%.[44] In all cases, it decays via the strong force. The pion channel would naïvely be the expected decay channel because the collective mass of the pions is smaller than the collective mass of the kaons, making the pion channel energetically favorable, but it channel is OZI suppressed, causing the decay distribution seen.

The ϕ meson can be thought of as a mix between the
s

s
,
u

u
, and
d

d
 states, but it is very nearly pure
s

s
.[45] This can be shown by deconstructing the wave function of the ϕ into its component parts. We see that the ϕ and ω are mixtures of the SU(3) wave functions as follows.

ϕ=ψ8cos θ - ψ1sin θ
ω=ψ8sin θ + ψ1cos θ

where

θ = the nonet mixing angle
ψ8=(u
u
 - d
d
 - 2s
s
)/√6
ψ1=(u
u
 - d
d
 + s
s
)/√3

The mixing angle at which the components decouple completely can be calculated to be about 35.3˚. The mixing angle of the ϕ and ω states is calculated from the masses of each state to be about 35˚, which is very close to maximum decoupling. We conclude that the ϕ phi meson is nearly pure
s

s
.[45]

History[edit]

The existence of the ϕ meson was first proposed by the Japanese American particle physicist, J. J. Sakurai, in 1962 as a resonance state between the
K0
 and the
K0
.[46] It was discovered later in 1962 by Connolly, et al. in a 20 inch hydrogen bubble chamber at the Alternating Gradient Synchrotron (AGS) in Brookhaven National Laboratory in Uptown, NY while they were studying
K−

p+
 collisions at approximately 2.23 GeV/c.[47] [48] In essence, the reaction involved a beam of
K−
s being accelerated to high energies to collide with protons.

The φ meson has several possible decay modes. The most energetically favored mode involves the φ meson decaying into 3 pions, which is what would naïvely be expected. However, we instead observe that it decays most frequently into 2 kaons.[49] Between 1963 and 1966, 3 people, Susumu Okubo, George Zweig and Jugoro Iizuka, each independently proposed a rule to account for the observed suppression of the 3 pion decay.[50] [51] [52] This rule is now known as the OZI rule and is also the currently accepted explanation for the unusually long lifetimes of the
J/ψ
 and
ϒ
 mesons.[49] Namely, on average they last ~ 7 × 10−21 s and ~ 1.5 × 10−20 s respectively.[49] This is compared to the normal mean lifetime of a meson decaying via the strong force, which is on the order of 10−23 s.[49]

In 1999, a new φ factory named DAFNE (or DAϕNE since the F stands for "ϕ Factory") began operation to study the decay of the φ meson in Frascati, Italy.[48] It produces φs via electron-positron collisions. It has numerous detectors, including the KLOE detector which was in operation at the beginning of it's operation.

References[edit]

  1. ^ citation
  2. ^ a b "Pythia". Papers With Code. Retrieved 1 July 2023.
  3. ^ a b c d e f g h i Biderman, Stella; Schoelkopf, Hailey; et al. (31 May 2023). "Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling". arXiv:2304.01373v2 [cs.CL].
  4. ^ a b Biderman, Stella; Schoelkopf, Hailey; Anthony, Quentin; Bradley, Herbie; O'Brien, Kyle; Hallahan, Eric; Khan, Mohammad Aflah; Purohit, Shivanshu; Prashanth, USVSN Sai; Raff, Edward; Skowron, Aviya; Sutawika, Lintang; van der Wal, Oskar (2023). "Pythia: Interpreting Transformers Across Time and Scale". GitHub. EleutherAI. Retrieved 1 July 2023.
  5. ^ Castelvecchi, Davide (5 October 2016). "Can we open the black box of AI?". Nature. Retrieved 2 July 2023. Unfortunately, such networks are also as opaque as the brain. Instead of storing what they have learned in a neat block of digital memory, they diffuse the information in a way that is exceedingly difficult to decipher.
  6. ^ Carlini, Nicholas; Tramèr, Florian; Wallace, Eric; Jagielski, Matthew; Herbert-Voss, Ariel; Katherine, Lee; Roberts, Adam; Brown, Tom; Song, Dawn; Erlingsson, Úlfar; Oprea, Alina; Raffel, Colin (11–13 August 2021). "Extracting Training Data from Large Language Models". Proceedings of the 30th USENIX Security Symposium. 30th USENIX Security Symposium (USENIX Security 21). USENIX Association. pp. 2633–2650. arXiv:2012.07805. ISBN 978-1-939133-24-3.
  7. ^ Biderman, Stella; Prashanth, USVSN Sai; Sutawika, Lintang; Schoelkopf, Hailey; Anthony, Quentin; Purohit, Shivanshu; Ruff, Edward (31 May 2023). "Emergent and Predictable Memorization in Large Language Models". arXiv:2304.11158v2 [cs.CL].
  8. ^ Razeghi, Yasaman; Logan, Robert L. IV; Gardner, Matt; Singh, Sameer (24 May 2022). "Impact of Pretraining Term Frequencies on Few-Shot Reasoning". arXiv:2202.07206 [cs.CL].
  9. ^ a b c d Leetaru, Kalev (28 September 2017). "Common Crawl And Unlocking Web Archives For Research". Forbes. Retrieved 9 June 2023.
  10. ^ a b "Common Crawl Foundation – Terms of Use". Common Crawl. 16 April 2010. Retrieved 8 June 2023.
  11. ^ Eskici (13 May 2023). "AI and Data Scraping on the Archive". Archive of Our Own. Retrieved 10 June 2023.
  12. ^ Leishman, Rachel (12 December 2022). "Fanfiction Writers Scramble To Set Profiles to Private as Evidence Grows That AI Writing Is Using Their Stories". The Mary Sue. Retrieved 10 June 2023.
  13. ^ "So you're ready to get started". Common Crawl. Retrieved 9 June 2023.
  14. ^ "How much memory do I need on my laptop?". Asurion. 2023. Retrieved 10 June 2023.
  15. ^ "17 U.S. Code § 512 - Limitations on liability relating to material online". Cornell Law School. 9 December 2010. Retrieved 9 June 2023. Title 17, US Code, Section 512(c)(3)(A)(iii)
  16. ^ Montti, Roger (2 February 2023). "How to Block ChatGPT From Using Your Website Content". Search Engine Journal. Retrieved 10 June 2023.
  17. ^ Montti, Roger (10 January 2023). "Is ChatGPT Use Of Web Content Fair?". Search Engine Journal. Retrieved 10 June 2023.
  18. ^ a b c Morin, Alexandre J. S., ed. (2017). Inclusive Physical Activities: International Perspectives. Information Age Publishing Inc. p. 294-297. ISBN 9781681238548.
  19. ^ a b c d e f g h i Pepper, David (September 25, 2007). Assessment for disabled students: an international comparison (PDF) (Report). UK: Ofqual's Qualifications and Curriculum Authority, Regulation & Standards Division. Archived (PDF) from the original on October 31, 2014.
  20. ^ a b c "Tokubetsu Shien Kyoiku". Archived from the original on 16 April 2021. Retrieved 22 November 2019.
  21. ^ a b Sasamoto, Ken (December 3–6, 2007). Final Report of the 27th Asia-Pacific International Seminar on Education for Individuals with Special Needs (PDF) (Report). Yokohama, Japan: National Institute of Special Needs Education (NISE). Retrieved July 23, 2023.
  22. ^ a b Tomes, Jason (April 13, 2010). "Some Thoughts on Special Education Classes in Japan's Junior High Schools". Retrieved July 22, 2023.
  23. ^ Minor, Kate (June 25, 2011). "なかよし　(Nakayoshi)". Blogspot. Retrieved July 22, 2023.
  24. ^ a b c d e f g h i Tsuge, T.; Munakata, T.; Tokunaga, Y.; Higo, S.; Takeda, T.; Osugi, N. (2002). "Special Support Education in Japan: Education for Children with Special Needs" (PDF). Ministry of Education, Culture, Sports, Science and Technology (MEXT). Retrieved July 21, 2023.
  25. ^ a b c d Mizutani, Yumi (July 2003). Individualized transition support plans for students with intellectual disabilities: implications for involvement of the students and the parents (Dissertation for a PhD in Education). University of Tsukuba. Retrieved July 20, 2023.
  26. ^ a b "Special Needs Education". Ministry of Education, Culture, Sports, Science and Technology (MEXT). Retrieved July 21, 2023.
  27. ^ 学校教育法施行規則 (in Japanese). Japanese government. 2007-12-25. Archived from the original on 2009-01-21. Retrieved 2008-02-16.
  28. ^ a b c d e f g h i j Yoneda, Hiroki (August 25, 2020). "Development of the Japanese Inclusive Education System: From Special Schools to Curriculum Modification for Special-Needs Education in Regular Schools". Proceedings of the 1st Progress in Social Science, Humanities and Education Research Symposium (PSSHERS 2019). 1st Progress in Social Science, Humanities and Education Research Symposium (PSSHERS 2019). Padang, West Sumatra, Indonesia: Atlantis Press. pp. 1031–1040. doi:10.2991/assehr.k.200824.228. ISSN 2352-5398.
  29. ^ Fang, Zuyi; Cerna-Turoff, Ilan; Zhang, Cheng; Lu, Mengyao; Lachman, Jamie M.; Barlow, Jane (May 2022). "Global estimates of violence against children with disabilities: an updated and systematic review and meta-analysis". Lancet Child Adolescent Health. 6 (5): 313–323. doi:10.1016/S2352-4642(22)00033-5. PMID 35305703. Retrieved July 23, 2023.
  30. ^ Horiguchi, Toshihiro; Takanashi, Kenji; Sato, Shoichi; Sone, Naoki (February 2, 2021). "Estimation of Abuse by Teachers in Special Needs Schools in Japan". Journal of Intellectual Disability Diagnosis and Treatment. 9 (1): 113–119. doi:10.6000/2292-2598.2021.09.01.14. Retrieved July 22, 2023.
  31. ^ a b c Shishido, Kazushige (May 2016). "About NISE". National Institute of Special Education (NISE). Retrieved July 22, 2023.
  32. ^ a b Gertz, Genie; Boudreault, Patrick (2016-01-05). The SAGE Deaf Studies Encyclopedia. SAGE Publications. ISBN 9781506341668.
  33. ^ Minami, Masahiko (2016-01-15). Handbook of Japanese Applied Linguistics. Walter de Gruyter GmbH & Co KG. ISBN 9781501500800.
  34. ^ Tsuge, Masayoshi. ""Teaching in Regular Classes" and "Pull-out Teaching": Substantiation of Each and Integration of the Two" (PDF). Hyōgo University of Teacher Education. Retrieved July 22, 2023. Before the middle ages, there were no "classes" in schools. / Then, "teaching by age" and "curricula by age" appeared (class system). /.../ In Japan, school classes were institutionalized in 1891 (Ordinance of the Ministry of Education, "Regulations on Organization of Classes").
  35. ^ "The UNESCO Salamanca Statement". Centre for Studies on Inclusive Education. November 18, 2020. Retrieved July 22, 2023.
  36. ^ a b United Nations General Assembly Session 61 Resolution 106. Convention on the Rights of Persons with Disabilities A/RES/61/106 December 13, 2006. Retrieved July 24, 2023. "alternate link: https://www.un.org/en/development/desa/population/migration/generalassembly/docs/globalcompact/A_RES_61_106.pdf"
  37. ^ a b c Furuta, Hiroko; Osugi, Nariki (September 30, 2016). "Developing an inclusive education system in Japan: The case of Yamaga City, Kumamoto" (PDF). Bulletin of the Faculty of Education, Kumamoto University. 65 (2016): 139–144. Retrieved July 21, 2023.
  38. ^ "Landmark UN treaty on rights of persons with disabilities enters into force". Scoop. 2008-05-05. Retrieved 2008-06-28.
  39. ^ a b c d "KLOE-2 | INFN-LNF". w3.lnf.infn.it. Retrieved 8 June 2017.
  40. ^ a b "The Nobel Prize in Physics 2008". www.nobelprize.org. Retrieved 8 June 2017.
  41. ^ a b Bossi, F. "Status of KLOE-2" (PDF). CERN. Retrieved 8 June 2017.
  42. ^ Franzini, Paolo; Moulson, Matthew (2006). "The Physics of DAFNE and KLOE". Annual Review of Nuclear and Particle Science. pp. 207–251. doi:10.1146/annurev.nucl.56.080805.140459. Retrieved 8 June 2017.
  43. ^ a b c d "Particle Detectors: KLOE" (PDF). INFN. Istituto Nazionale di Fisica Nucleare. Retrieved 8 June 2017.
  44. ^ a b c Nakamura, K.; et al. "Particle listings – φ" (PDF). Retrieved 5 May 2017. {{cite web}}: Explicit use of et al. in: |first1= (help)
  45. ^ a b Nakamura, K. "14. Quark Model" (PDF). Retrieved 5 May 2017.
  46. ^ Sakurai, J. J. (1 December 1962). "Possible Existence of a T=0 Vector Meson at 1020 MeV". Physical Review Letters. pp. 472–475. doi:10.1103/PhysRevLett.9.472. Retrieved 5 May 2017.
  47. ^ Connolly, P. L.; Hart, E. L.; Lai, K. W.; London, G.; Moneti, G. C.; Rau, R. R.; Samios, N. P.; Skillicorn, I. O.; Yamamoto, S. S.; Goldberg, M.; Gundzik, M.; Leitner, J.; Lichtman, S. (15 April 1963). "Existence and Properties of the ϕ Meson". Physical Review Letters. pp. 371–376. doi:10.1103/PhysRevLett.10.371. Retrieved 5 May 2017. {{cite web}}: line feed character in |title= at position 34 (help)
  48. ^ a b "K for KLOE... ...and Z for Zweig - CERN Courier". cerncourier.com. Retrieved 6 May 2017.
  49. ^ a b c d Griffiths, David (2008). Introduction to elementary particles (2nd rev. ed. ed.). Weinheim: Wiley-VCH. ISBN 978-3-527-40601-2. {{cite book}}: |edition= has extra text (help)
  50. ^ S. Okubo, Phys. Lett. 5, 1975 (1963).
  51. ^ G. Zweig, CERN Report No.8419/TH412 (1964).
  52. ^ J. Iizuka, Prog. Theor. Phys. Suppl. 37, 21 (1966).
Retrieved from "https://en.wikipedia.org/w/index.php?title=User:Penrose_Delta/sandbox&oldid=1173505651"