- published: 16 Feb 2023
- views: 2083
-
remove the playlistInteractive Theorem Proving (conference)
- remove the playlistInteractive Theorem Proving (conference)
Interactive Theorem Proving (conference)
Interactive Theorem Proving (ITP) is an annual international academic conference on the topic of automated theorem proving, proof assistants and related topics, ranging from theoretical foundations to implementation aspects and applications in program verification, security, and formalization of mathematics.
ITP brings together the communities using many systems based on higher-order logic such as ACL2, Coq, Mizar, HOL, Isabelle, NuPRL, PVS, and Twelf. Individual workshops or meetings devoted to individual systems are usually held concurrently with the conference.
Together with CADE and TABLEAUX, ITP is usually one of the three main conferences of the International Joint Conference on Automated Reasoning (IJCAR) whenever it convenes,
History
The inaugural meeting of ITP was held on 11–14 July 2010 in Edinburgh, Scotland, as part of the Federated Logic Conference. It is the extension of the Theorem Proving in Higher Order Logics (TPHOLs) conference series to the broad field of interactive theorem proving. TPHOLs meetings took place every year from 1988 until 2009.
This page contains text from Wikipedia, the Free Encyclopedia - https://wn.com/Interactive_Theorem_Proving_(conference)
Proof assistant
In computer science and mathematical logic, a proof assistant or interactive theorem prover is a software tool to assist with the development of formal proofs by human-machine collaboration. This involves some sort of interactive proof editor, or other interface, with which a human can guide the search for proofs, the details of which are stored in, and some steps provided by, a computer.
Comparison of systems
- HOL4 – The "primary descendant", still under active development. Support for both Moscow ML and Poly/ML. Has a BSD-style license.
- HOL Light – A thriving "minimalist fork". OCaml based.
- ProofPower – Went proprietary, then returned to open source. Based on Standard ML.
This page contains text from Wikipedia, the Free Encyclopedia - https://wn.com/Proof_assistant
Podcasts:
-
Jason Rute - Deep learning in interactive theorem proving - IPAM at UCLA
Recorded 16 February 2023. Jason Rute of IBM presents "Deep learning in interactive theorem proving" at IPAM's Machine Assisted Proofs Workshop. Abstract: Deep learning has made progress in many diverse areas, often leveraging a relatively small toolbox of powerful techniques. One promising area of application is formal mathematics and interactive theorem proving. I will talk about past progress and future possibilities, focusing on the big picture. I will also talk about the many challenges and how they connect to challenges in machine learning as a whole. Learn more online at: http://www.ipam.ucla.edu/programs/workshops/machine-assisted-proofs/
published: 16 Feb 2023 -
Interactive Theorem Proving with Lean
Cody Roux New York Haskell Meetup (http://www.meetup.com/NY-Haskell/events/226667224/) November 25, 2015 Slides: https://github.com/codyroux/ny-haskell/blob/master/slides/talk.pdf
published: 01 Dec 2015 -
Towards Lean 4: An Optimized Object Model for an Interactive Theorem Prover
http://www.pollylabs.org/llvm-social-zurich.html Lean 4, the next version of the Lean theorem prover, will move most of its frontend code from C++ to Lean itself. To ensure that the resulting code is reasonably efficient, Lean 4 will feature a new code generation backend together with a new object and memory management model. In this talk, I will discuss the general and ITP-specific constraints, such as performance, language interoperability, and startup time, that led us to this model and how we are planning to solve them with it. Speaker: Sebastian Ullrich is a second-year PhD student at Gregor Snelting's programming paradigms group at the Karlsruhe Institute of Technology (KIT). He is working on the Lean theorem prover together with Leonardo de Moura (Microsoft Research). Sebastian's...
published: 18 Dec 2018 -
CAIS-23-03 | Professor Lawrence Paulson | Automated Theorem Proving: A Technology Roadmap
About CAIS [https://www.cambridgeaisocial.org] Cambridge AI Social (CAIS) seeks to deliver a series of in-person “AI + pizza” events (CAIS Lectures) in Cambridge, UK Fundamental to the CAIS vision are: - speakers must be a recognised world leader in their field - events are non-profit, with minimal cost to attendees - each event comprises a roughly one-hour talk followed by socialisation (with pizza!) About the talk Title: “Automated Theorem Proving: a Technology Roadmap” Abstract: The technology of automated deduction has a long pedigree. For ordinary first-order logic, the basic techniques had all been invented by 1965: DPLL (for large Boolean problems) and the tableau and resolution calculi (for quantifiers). The relationship between automated deduction and AI has been complex: do...
published: 10 Jun 2023 -
Non Interactive Zero Knowledge Proofs for Composite Statements
Paper by Shashank Agrawal and Chaya Ganesh and Payman Mohassel, presented at Crypto 2018. See https://iacr.org/cryptodb/data/paper.php?pubkey=28802
published: 05 Oct 2018 -
Proving Theorems & Seeing Cats
Presented at SPLASH-I 2018 First off, you’re not going to learn anything about programming techniques, software development, or really, anything useful from this talk. I will tell you how a program I wrote for DARPA to help thwart terrorist plots turned into one that could write the following “haiku” when inspired by the phrase “loud guitar blues music”: tuned adrenalin my music a beat-boogied headful Along the way I will tell you two things: how and why programming is stranger than you think why real artificial intelligence needs to be a blend of proving theorems (in deference to John McCarthy) and seeing cats (in deference to Andrew Ng) Geoffrey Jefferson, more or less debating Alan Turing, wrote: “Not until a machine can write a sonnet or compose a concerto because of thoughts ...
published: 21 Nov 2018 -
Josef Urban | AI and Theorem Proving
1/13/2021 New Technologies in Mathematics Seminar Speaker: Josef Urban, Czech Institute of Informatics, Robotics and Cybernetics Title: AI and Theorem Proving Abstract: The talk will discuss the main approaches that combine machine learning with automated theorem proving and automated formalization. This includes learning to choose relevant facts for “hammer” systems, guiding the proof search of tableaux and superposition automated provers by interleaving learning and proving (reinforcement learning) over large ITP libraries, guiding the application of tactics in interactive tactical systems, and various forms of lemmatization and conjecturing. I will also show some demos of the systems, and discuss autoformalization approaches such as learning probabilistic grammars from aligned inform...
published: 19 Jan 2021 -
20190605 Introduction to Interactive theorem proving, at the OSU Quantum Symmetries summer school
published: 05 Jun 2019 -
Kevin Buzzard: Mathematics and the Computer with G-Research
Kevin Buzzard, Professor of Pure Mathematics at Imperial College London, discusses Mathematics and the Computer as part of G-Research's Computer Guided Mathematics Symposium. Held at the Science Museum in Central London, Kevin was the first of three speakers, alongside Alex Davies (DeepMind) and Sir Timothy Gowers (Professeur titulaire of the Combinatorics chair at the Collège de France). Want to watch the talks? Check out our Computer Guided Mathematics Symposium playlist here: https://www.youtube.com/playlist?list=PL_9Cdw79RQrS65GRmrYaauOOygLpXdmMv View Kevin's full presentation slides: http://bit.ly/3CHKYbc The G-Research Computer Guided Mathematics Symposium was the final event in our 2022 Distinguished Speaker Series. We'll be running events in 2023, so if you're interested in at...
published: 22 Dec 2022 -
Introducing PUTNAMBENCH: Revolutionizing AI Theorem Proving
UT Austin Researchers Launch PUTNAMBENCH: A Milestone in AI Benchmarking Researchers at the University of Texas at Austin have unveiled PUTNAMBENCH, an innovative benchmark designed to evaluate neural theorem-provers by utilizing problems from the highly competitive William Lowell Putnam Mathematical Competition. This benchmark addresses a critical gap in AI development, where existing benchmarks focus too much on high-school-level mathematics and fail to challenge theorem-provers with advanced problems. PUTNAMBENCH offers 1697 formalizations of 640 problems available in multiple formal proof languages, including Lean 4, Isabelle, and Coq. The meticulous creation of these problems aims to ensure rigorous testing across different theorem-proving environments. Despite current methods like...
published: 20 Jul 2024
developed with YouTube
55:28
Jason Rute - Deep learning in interactive theorem proving - IPAM at UCLA
- Order: Reorder
- Duration: 55:28
- Uploaded Date: 16 Feb 2023
- views: 2083
Recorded 16 February 2023. Jason Rute of IBM presents "Deep learning in interactive theorem proving" at IPAM's Machine Assisted Proofs Workshop.
Abstract: Deep ...
Recorded 16 February 2023. Jason Rute of IBM presents "Deep learning in interactive theorem proving" at IPAM's Machine Assisted Proofs Workshop.
Abstract: Deep learning has made progress in many diverse areas, often leveraging a relatively small toolbox of powerful techniques. One promising area of application is formal mathematics and interactive theorem proving. I will talk about past progress and future possibilities, focusing on the big picture. I will also talk about the many challenges and how they connect to challenges in machine learning as a whole.
Learn more online at: http://www.ipam.ucla.edu/programs/workshops/machine-assisted-proofs/
https://wn.com/Jason_Rute_Deep_Learning_In_Interactive_Theorem_Proving_Ipam_At_Ucla
Recorded 16 February 2023. Jason Rute of IBM presents "Deep learning in interactive theorem proving" at IPAM's Machine Assisted Proofs Workshop.
Abstract: Deep learning has made progress in many diverse areas, often leveraging a relatively small toolbox of powerful techniques. One promising area of application is formal mathematics and interactive theorem proving. I will talk about past progress and future possibilities, focusing on the big picture. I will also talk about the many challenges and how they connect to challenges in machine learning as a whole.
Learn more online at: http://www.ipam.ucla.edu/programs/workshops/machine-assisted-proofs/
1:49:14
Interactive Theorem Proving with Lean
- Order: Reorder
- Duration: 1:49:14
- Uploaded Date: 01 Dec 2015
- views: 4638
Cody Roux
New York Haskell Meetup (http://www.meetup.com/NY-Haskell/events/226667224/)
November 25, 2015
Slides: https://github.com/codyroux/ny-haskell/blob/ma...
Cody Roux
New York Haskell Meetup (http://www.meetup.com/NY-Haskell/events/226667224/)
November 25, 2015
Slides: https://github.com/codyroux/ny-haskell/blob/master/slides/talk.pdf
https://wn.com/Interactive_Theorem_Proving_With_Lean
Cody Roux
New York Haskell Meetup (http://www.meetup.com/NY-Haskell/events/226667224/)
November 25, 2015
Slides: https://github.com/codyroux/ny-haskell/blob/master/slides/talk.pdf
- published: 01 Dec 2015
- views: 4638
48:02
Towards Lean 4: An Optimized Object Model for an Interactive Theorem Prover
- Order: Reorder
- Duration: 48:02
- Uploaded Date: 18 Dec 2018
- views: 1341
http://www.pollylabs.org/llvm-social-zurich.html
Lean 4, the next version of the Lean theorem prover, will move most of its frontend code from C++ to Lean itse...
http://www.pollylabs.org/llvm-social-zurich.html
Lean 4, the next version of the Lean theorem prover, will move most of its frontend code from C++ to Lean itself. To ensure that the resulting code is reasonably efficient, Lean 4 will feature a new code generation backend together with a new object and memory management model. In this talk, I will discuss the general and ITP-specific constraints, such as performance, language interoperability, and startup time, that led us to this model and how we are planning to solve them with it.
Speaker:
Sebastian Ullrich is a second-year PhD student at Gregor Snelting's programming paradigms group at the Karlsruhe Institute of Technology (KIT). He is working on the Lean theorem prover together with Leonardo de Moura (Microsoft Research). Sebastian's research interests are program verification, the design of interactive theorem proving frontends, and macro expansion.
https://wn.com/Towards_Lean_4_An_Optimized_Object_Model_For_An_Interactive_Theorem_Prover
http://www.pollylabs.org/llvm-social-zurich.html
Lean 4, the next version of the Lean theorem prover, will move most of its frontend code from C++ to Lean itself. To ensure that the resulting code is reasonably efficient, Lean 4 will feature a new code generation backend together with a new object and memory management model. In this talk, I will discuss the general and ITP-specific constraints, such as performance, language interoperability, and startup time, that led us to this model and how we are planning to solve them with it.
Speaker:
Sebastian Ullrich is a second-year PhD student at Gregor Snelting's programming paradigms group at the Karlsruhe Institute of Technology (KIT). He is working on the Lean theorem prover together with Leonardo de Moura (Microsoft Research). Sebastian's research interests are program verification, the design of interactive theorem proving frontends, and macro expansion.
- published: 18 Dec 2018
- views: 1341
59:55
CAIS-23-03 | Professor Lawrence Paulson | Automated Theorem Proving: A Technology Roadmap
- Order: Reorder
- Duration: 59:55
- Uploaded Date: 10 Jun 2023
- views: 230
About CAIS [https://www.cambridgeaisocial.org]
Cambridge AI Social (CAIS) seeks to deliver a series of in-person “AI + pizza” events (CAIS Lectures) in Cambrid...
About CAIS [https://www.cambridgeaisocial.org]
Cambridge AI Social (CAIS) seeks to deliver a series of in-person “AI + pizza” events (CAIS Lectures) in Cambridge, UK
Fundamental to the CAIS vision are:
- speakers must be a recognised world leader in their field
- events are non-profit, with minimal cost to attendees
- each event comprises a roughly one-hour talk followed by socialisation (with pizza!)
About the talk
Title: “Automated Theorem Proving: a Technology Roadmap”
Abstract: The technology of automated deduction has a long pedigree. For ordinary first-order logic, the basic techniques had all been invented by 1965: DPLL (for large Boolean problems) and the tableau and resolution calculi (for quantifiers). The relationship between automated deduction and AI has been complex: does intelligence emerge from deduction, or is it the other way around? Interactive theorem proving further complicates the picture, with a human user working in a formal calculus much stronger than first-order logic on huge, open-ended verification problems and needing maximum automation. Isabelle is an example of a sophisticated interactive prover that also relies heavily on automatic technologies through its nitpick and sledgehammer subsystems. The talk will give an architectural overview of Isabelle and its associated tools. The speaker will also speculate on how future developments, especially machine learning, could assist (not replace) the user.
About the speaker
Paulson graduated from the California Institute of Technology with a BS in Mathematics in 1977, and obtained a PhD in Computer Science from Stanford University in 1981. After a brief spell as a research assistant at Edinburgh University, Paulson moved to Cambridge University in 1983, where he has been ever since. Elevated to Professor of Computational Logic in 2002, a position he held for 20 years, Paulson has been Director of Research at the University of Cambridge Computer Laboratory since 2022. Paulson was elected Fellow of the ACM (Association of Computing Machinery) in 2008, and Fellow of the Royal Society in 2017. He has been an editor of the Journal of Automated Reasoning for many years, and a trustee of the Conference on Automated Deduction (CADE) since 2010.
https://wn.com/Cais_23_03_|_Professor_Lawrence_Paulson_|_Automated_Theorem_Proving_A_Technology_Roadmap
About CAIS [https://www.cambridgeaisocial.org]
Cambridge AI Social (CAIS) seeks to deliver a series of in-person “AI + pizza” events (CAIS Lectures) in Cambridge, UK
Fundamental to the CAIS vision are:
- speakers must be a recognised world leader in their field
- events are non-profit, with minimal cost to attendees
- each event comprises a roughly one-hour talk followed by socialisation (with pizza!)
About the talk
Title: “Automated Theorem Proving: a Technology Roadmap”
Abstract: The technology of automated deduction has a long pedigree. For ordinary first-order logic, the basic techniques had all been invented by 1965: DPLL (for large Boolean problems) and the tableau and resolution calculi (for quantifiers). The relationship between automated deduction and AI has been complex: does intelligence emerge from deduction, or is it the other way around? Interactive theorem proving further complicates the picture, with a human user working in a formal calculus much stronger than first-order logic on huge, open-ended verification problems and needing maximum automation. Isabelle is an example of a sophisticated interactive prover that also relies heavily on automatic technologies through its nitpick and sledgehammer subsystems. The talk will give an architectural overview of Isabelle and its associated tools. The speaker will also speculate on how future developments, especially machine learning, could assist (not replace) the user.
About the speaker
Paulson graduated from the California Institute of Technology with a BS in Mathematics in 1977, and obtained a PhD in Computer Science from Stanford University in 1981. After a brief spell as a research assistant at Edinburgh University, Paulson moved to Cambridge University in 1983, where he has been ever since. Elevated to Professor of Computational Logic in 2002, a position he held for 20 years, Paulson has been Director of Research at the University of Cambridge Computer Laboratory since 2022. Paulson was elected Fellow of the ACM (Association of Computing Machinery) in 2008, and Fellow of the Royal Society in 2017. He has been an editor of the Journal of Automated Reasoning for many years, and a trustee of the Conference on Automated Deduction (CADE) since 2010.
- published: 10 Jun 2023
- views: 230
21:02
Non Interactive Zero Knowledge Proofs for Composite Statements
- Order: Reorder
- Duration: 21:02
- Uploaded Date: 05 Oct 2018
- views: 1821
Paper by Shashank Agrawal and Chaya Ganesh and Payman Mohassel, presented at Crypto 2018. See https://iacr.org/cryptodb/data/paper.php?pubkey=28802
Paper by Shashank Agrawal and Chaya Ganesh and Payman Mohassel, presented at Crypto 2018. See https://iacr.org/cryptodb/data/paper.php?pubkey=28802
https://wn.com/Non_Interactive_Zero_Knowledge_Proofs_For_Composite_Statements
Paper by Shashank Agrawal and Chaya Ganesh and Payman Mohassel, presented at Crypto 2018. See https://iacr.org/cryptodb/data/paper.php?pubkey=28802
- published: 05 Oct 2018
- views: 1821
1:07:51
Proving Theorems & Seeing Cats
- Order: Reorder
- Duration: 1:07:51
- Uploaded Date: 21 Nov 2018
- views: 100
Presented at SPLASH-I 2018
First off, you’re not going to learn anything about programming techniques, software development, or really, anything useful from th...
Presented at SPLASH-I 2018
First off, you’re not going to learn anything about programming techniques, software development, or really, anything useful from this talk. I will tell you how a program I wrote for DARPA to help thwart terrorist plots turned into one that could write the following “haiku” when inspired by the phrase “loud guitar blues music”:
tuned adrenalin
my music
a beat-boogied headful
Along the way I will tell you two things:
how and why programming is stranger than you think
why real artificial intelligence needs to be a blend of proving theorems (in deference to John McCarthy) and seeing cats (in deference to Andrew Ng)
Geoffrey Jefferson, more or less debating Alan Turing, wrote: “Not until a machine can write a sonnet or compose a concerto because of thoughts and emotions felt, and not by the chance fall of symbols, could we agree that machine equals brain—that is, not only write it but know that it had written it.” [‘The Mind of Mechanical Man,’ British Medical Journal, 1949]
Stopping by Sums on a Snowy Mean Solar Day by the Roberts Frost and Harper
Whose sums these are I express I know.
His unit is in the body though;
He will not see me entering existing
To consider his sums operate up with snow.
My smaller horse must evaluate it interesting
To keep in without a family discriminating
Between the sums and polar form
The darkest mean solar day of the fundamental quantity.
He gives his influence devices an exit
To pass along if there is some statement.
The only other element is the use
Of benign context and yielding element.
The sums are logical, insensitive, and severe.
But I have expressions to position,
And large indefinite amounts to go before I sleep in,
And large indefinite amounts to go before I position.
https://wn.com/Proving_Theorems_Seeing_Cats
Presented at SPLASH-I 2018
First off, you’re not going to learn anything about programming techniques, software development, or really, anything useful from this talk. I will tell you how a program I wrote for DARPA to help thwart terrorist plots turned into one that could write the following “haiku” when inspired by the phrase “loud guitar blues music”:
tuned adrenalin
my music
a beat-boogied headful
Along the way I will tell you two things:
how and why programming is stranger than you think
why real artificial intelligence needs to be a blend of proving theorems (in deference to John McCarthy) and seeing cats (in deference to Andrew Ng)
Geoffrey Jefferson, more or less debating Alan Turing, wrote: “Not until a machine can write a sonnet or compose a concerto because of thoughts and emotions felt, and not by the chance fall of symbols, could we agree that machine equals brain—that is, not only write it but know that it had written it.” [‘The Mind of Mechanical Man,’ British Medical Journal, 1949]
Stopping by Sums on a Snowy Mean Solar Day by the Roberts Frost and Harper
Whose sums these are I express I know.
His unit is in the body though;
He will not see me entering existing
To consider his sums operate up with snow.
My smaller horse must evaluate it interesting
To keep in without a family discriminating
Between the sums and polar form
The darkest mean solar day of the fundamental quantity.
He gives his influence devices an exit
To pass along if there is some statement.
The only other element is the use
Of benign context and yielding element.
The sums are logical, insensitive, and severe.
But I have expressions to position,
And large indefinite amounts to go before I sleep in,
And large indefinite amounts to go before I position.
- published: 21 Nov 2018
- views: 100
1:22:59
Josef Urban | AI and Theorem Proving
- Order: Reorder
- Duration: 1:22:59
- Uploaded Date: 19 Jan 2021
- views: 3063
1/13/2021 New Technologies in Mathematics Seminar
Speaker: Josef Urban, Czech Institute of Informatics, Robotics and Cybernetics
Title: AI and Theorem Proving...
1/13/2021 New Technologies in Mathematics Seminar
Speaker: Josef Urban, Czech Institute of Informatics, Robotics and Cybernetics
Title: AI and Theorem Proving
Abstract: The talk will discuss the main approaches that combine machine learning with automated theorem proving and automated formalization. This includes learning to choose relevant facts for “hammer” systems, guiding the proof search of tableaux and superposition automated provers by interleaving learning and proving (reinforcement learning) over large ITP libraries, guiding the application of tactics in interactive tactical systems, and various forms of lemmatization and conjecturing. I will also show some demos of the systems, and discuss autoformalization approaches such as learning probabilistic grammars from aligned informal/formal corpora, combining them with semantic pruning, and using neural methods to learn direct translation from Latex to formal mathematics.
https://wn.com/Josef_Urban_|_Ai_And_Theorem_Proving
1/13/2021 New Technologies in Mathematics Seminar
Speaker: Josef Urban, Czech Institute of Informatics, Robotics and Cybernetics
Title: AI and Theorem Proving
Abstract: The talk will discuss the main approaches that combine machine learning with automated theorem proving and automated formalization. This includes learning to choose relevant facts for “hammer” systems, guiding the proof search of tableaux and superposition automated provers by interleaving learning and proving (reinforcement learning) over large ITP libraries, guiding the application of tactics in interactive tactical systems, and various forms of lemmatization and conjecturing. I will also show some demos of the systems, and discuss autoformalization approaches such as learning probabilistic grammars from aligned informal/formal corpora, combining them with semantic pruning, and using neural methods to learn direct translation from Latex to formal mathematics.
- published: 19 Jan 2021
- views: 3063
1:09:54
20190605 Introduction to Interactive theorem proving, at the OSU Quantum Symmetries summer school
- Order: Reorder
- Duration: 1:09:54
- Uploaded Date: 05 Jun 2019
- views: 1299
- published: 05 Jun 2019
- views: 1299
43:54
Kevin Buzzard: Mathematics and the Computer with G-Research
- Order: Reorder
- Duration: 43:54
- Uploaded Date: 22 Dec 2022
- views: 3485
Kevin Buzzard, Professor of Pure Mathematics at Imperial College London, discusses Mathematics and the Computer as part of G-Research's Computer Guided Mathemat...
Kevin Buzzard, Professor of Pure Mathematics at Imperial College London, discusses Mathematics and the Computer as part of G-Research's Computer Guided Mathematics Symposium.
Held at the Science Museum in Central London, Kevin was the first of three speakers, alongside Alex Davies (DeepMind) and Sir Timothy Gowers (Professeur titulaire of the Combinatorics chair at the Collège de France).
Want to watch the talks? Check out our Computer Guided Mathematics Symposium playlist here: https://www.youtube.com/playlist?list=PL_9Cdw79RQrS65GRmrYaauOOygLpXdmMv
View Kevin's full presentation slides: http://bit.ly/3CHKYbc
The G-Research Computer Guided Mathematics Symposium was the final event in our 2022 Distinguished Speaker Series. We'll be running events in 2023, so if you're interested in attending future Distinguished Speaker Series events, please register your details here: https://events.beamery.com/gresearch/all-dss-events-mntauiaxr
https://wn.com/Kevin_Buzzard_Mathematics_And_The_Computer_With_G_Research
Kevin Buzzard, Professor of Pure Mathematics at Imperial College London, discusses Mathematics and the Computer as part of G-Research's Computer Guided Mathematics Symposium.
Held at the Science Museum in Central London, Kevin was the first of three speakers, alongside Alex Davies (DeepMind) and Sir Timothy Gowers (Professeur titulaire of the Combinatorics chair at the Collège de France).
Want to watch the talks? Check out our Computer Guided Mathematics Symposium playlist here: https://www.youtube.com/playlist?list=PL_9Cdw79RQrS65GRmrYaauOOygLpXdmMv
View Kevin's full presentation slides: http://bit.ly/3CHKYbc
The G-Research Computer Guided Mathematics Symposium was the final event in our 2022 Distinguished Speaker Series. We'll be running events in 2023, so if you're interested in attending future Distinguished Speaker Series events, please register your details here: https://events.beamery.com/gresearch/all-dss-events-mntauiaxr
- published: 22 Dec 2022
- views: 3485
5:04
Introducing PUTNAMBENCH: Revolutionizing AI Theorem Proving
- Order: Reorder
- Duration: 5:04
- Uploaded Date: 20 Jul 2024
- views: 37
UT Austin Researchers Launch PUTNAMBENCH: A Milestone in AI Benchmarking
Researchers at the University of Texas at Austin have unveiled PUTNAMBENCH, an innovat...
UT Austin Researchers Launch PUTNAMBENCH: A Milestone in AI Benchmarking
Researchers at the University of Texas at Austin have unveiled PUTNAMBENCH, an innovative benchmark designed to evaluate neural theorem-provers by utilizing problems from the highly competitive William Lowell Putnam Mathematical Competition. This benchmark addresses a critical gap in AI development, where existing benchmarks focus too much on high-school-level mathematics and fail to challenge theorem-provers with advanced problems.
PUTNAMBENCH offers 1697 formalizations of 640 problems available in multiple formal proof languages, including Lean 4, Isabelle, and Coq. The meticulous creation of these problems aims to ensure rigorous testing across different theorem-proving environments. Despite current methods like GPT-4 and Sledgehammer being tested, the results indicate that only a few problems were successfully solved. This highlights the ongoing challenges faced in automating mathematical reasoning
With PUTNAMBENCH, researchers hope to drive future innovation and improvements in neural theorem provers. What are your thoughts on the impact of advanced benchmarks like PUTNAMBENCH on the future of AI in mathematical problem-solving?
Explore how Klap AI can transform your videos into viral short content at www.TheBestAI.org/now.
#AI #research #mathematics #innovation
Unveiling PUTNAMBENCH: The Ultimate AI Benchmark! by Steven's Workspace
OUTLINE:
00:00:00 Introducing PUTNAMBENCH
00:00:49 Automating Mathematical Reasoning
00:01:23 The Need for a Challenging Benchmark
00:02:37 The Genesis of PUTNAMBENCH
00:03:41 Multiple Formalisms in PUTNAMBENCH
00:04:51 The Potential of PUTNAMBENCH
https://wn.com/Introducing_Putnambench_Revolutionizing_Ai_Theorem_Proving
UT Austin Researchers Launch PUTNAMBENCH: A Milestone in AI Benchmarking
Researchers at the University of Texas at Austin have unveiled PUTNAMBENCH, an innovative benchmark designed to evaluate neural theorem-provers by utilizing problems from the highly competitive William Lowell Putnam Mathematical Competition. This benchmark addresses a critical gap in AI development, where existing benchmarks focus too much on high-school-level mathematics and fail to challenge theorem-provers with advanced problems.
PUTNAMBENCH offers 1697 formalizations of 640 problems available in multiple formal proof languages, including Lean 4, Isabelle, and Coq. The meticulous creation of these problems aims to ensure rigorous testing across different theorem-proving environments. Despite current methods like GPT-4 and Sledgehammer being tested, the results indicate that only a few problems were successfully solved. This highlights the ongoing challenges faced in automating mathematical reasoning
With PUTNAMBENCH, researchers hope to drive future innovation and improvements in neural theorem provers. What are your thoughts on the impact of advanced benchmarks like PUTNAMBENCH on the future of AI in mathematical problem-solving?
Explore how Klap AI can transform your videos into viral short content at www.TheBestAI.org/now.
#AI #research #mathematics #innovation
Unveiling PUTNAMBENCH: The Ultimate AI Benchmark! by Steven's Workspace
OUTLINE:
00:00:00 Introducing PUTNAMBENCH
00:00:49 Automating Mathematical Reasoning
00:01:23 The Need for a Challenging Benchmark
00:02:37 The Genesis of PUTNAMBENCH
00:03:41 Multiple Formalisms in PUTNAMBENCH
00:04:51 The Potential of PUTNAMBENCH
- published: 20 Jul 2024
- views: 37
55:28
Jason Rute - Deep learning in interactive theorem proving - IPAM at UCLA
Recorded 16 February 2023. Jason Rute of IBM presents "Deep learning in interactive theore...
published: 16 Feb 2023
Jason Rute - Deep learning in interactive theorem proving - IPAM at UCLA
Jason Rute - Deep learning in interactive theorem proving - IPAM at UCLA
- Report rights infringement
- published: 16 Feb 2023
- views: 2083
Recorded 16 February 2023. Jason Rute of IBM presents "Deep learning in interactive theorem proving" at IPAM's Machine Assisted Proofs Workshop.
Abstract: Deep learning has made progress in many diverse areas, often leveraging a relatively small toolbox of powerful techniques. One promising area of application is formal mathematics and interactive theorem proving. I will talk about past progress and future possibilities, focusing on the big picture. I will also talk about the many challenges and how they connect to challenges in machine learning as a whole.
Learn more online at: http://www.ipam.ucla.edu/programs/workshops/machine-assisted-proofs/
1:49:14
Interactive Theorem Proving with Lean
Cody Roux
New York Haskell Meetup (http://www.meetup.com/NY-Haskell/events/226667224/)
Nov...
published: 01 Dec 2015
Interactive Theorem Proving with Lean
Interactive Theorem Proving with Lean
- Report rights infringement
- published: 01 Dec 2015
- views: 4638
Cody Roux
New York Haskell Meetup (http://www.meetup.com/NY-Haskell/events/226667224/)
November 25, 2015
Slides: https://github.com/codyroux/ny-haskell/blob/master/slides/talk.pdf
48:02
Towards Lean 4: An Optimized Object Model for an Interactive Theorem Prover
http://www.pollylabs.org/llvm-social-zurich.html
Lean 4, the next version of the Lean the...
published: 18 Dec 2018
Towards Lean 4: An Optimized Object Model for an Interactive Theorem Prover
Towards Lean 4: An Optimized Object Model for an Interactive Theorem Prover
- Report rights infringement
- published: 18 Dec 2018
- views: 1341
http://www.pollylabs.org/llvm-social-zurich.html
Lean 4, the next version of the Lean theorem prover, will move most of its frontend code from C++ to Lean itself. To ensure that the resulting code is reasonably efficient, Lean 4 will feature a new code generation backend together with a new object and memory management model. In this talk, I will discuss the general and ITP-specific constraints, such as performance, language interoperability, and startup time, that led us to this model and how we are planning to solve them with it.
Speaker:
Sebastian Ullrich is a second-year PhD student at Gregor Snelting's programming paradigms group at the Karlsruhe Institute of Technology (KIT). He is working on the Lean theorem prover together with Leonardo de Moura (Microsoft Research). Sebastian's research interests are program verification, the design of interactive theorem proving frontends, and macro expansion.
59:55
CAIS-23-03 | Professor Lawrence Paulson | Automated Theorem Proving: A Technology Roadmap
About CAIS [https://www.cambridgeaisocial.org]
Cambridge AI Social (CAIS) seeks to delive...
published: 10 Jun 2023
CAIS-23-03 | Professor Lawrence Paulson | Automated Theorem Proving: A Technology Roadmap
CAIS-23-03 | Professor Lawrence Paulson | Automated Theorem Proving: A Technology Roadmap
- Report rights infringement
- published: 10 Jun 2023
- views: 230
About CAIS [https://www.cambridgeaisocial.org]
Cambridge AI Social (CAIS) seeks to deliver a series of in-person “AI + pizza” events (CAIS Lectures) in Cambridge, UK
Fundamental to the CAIS vision are:
- speakers must be a recognised world leader in their field
- events are non-profit, with minimal cost to attendees
- each event comprises a roughly one-hour talk followed by socialisation (with pizza!)
About the talk
Title: “Automated Theorem Proving: a Technology Roadmap”
Abstract: The technology of automated deduction has a long pedigree. For ordinary first-order logic, the basic techniques had all been invented by 1965: DPLL (for large Boolean problems) and the tableau and resolution calculi (for quantifiers). The relationship between automated deduction and AI has been complex: does intelligence emerge from deduction, or is it the other way around? Interactive theorem proving further complicates the picture, with a human user working in a formal calculus much stronger than first-order logic on huge, open-ended verification problems and needing maximum automation. Isabelle is an example of a sophisticated interactive prover that also relies heavily on automatic technologies through its nitpick and sledgehammer subsystems. The talk will give an architectural overview of Isabelle and its associated tools. The speaker will also speculate on how future developments, especially machine learning, could assist (not replace) the user.
About the speaker
Paulson graduated from the California Institute of Technology with a BS in Mathematics in 1977, and obtained a PhD in Computer Science from Stanford University in 1981. After a brief spell as a research assistant at Edinburgh University, Paulson moved to Cambridge University in 1983, where he has been ever since. Elevated to Professor of Computational Logic in 2002, a position he held for 20 years, Paulson has been Director of Research at the University of Cambridge Computer Laboratory since 2022. Paulson was elected Fellow of the ACM (Association of Computing Machinery) in 2008, and Fellow of the Royal Society in 2017. He has been an editor of the Journal of Automated Reasoning for many years, and a trustee of the Conference on Automated Deduction (CADE) since 2010.
21:02
Non Interactive Zero Knowledge Proofs for Composite Statements
Paper by Shashank Agrawal and Chaya Ganesh and Payman Mohassel, presented at Crypto 2018. ...
published: 05 Oct 2018
Non Interactive Zero Knowledge Proofs for Composite Statements
Non Interactive Zero Knowledge Proofs for Composite Statements
- Report rights infringement
- published: 05 Oct 2018
- views: 1821
Paper by Shashank Agrawal and Chaya Ganesh and Payman Mohassel, presented at Crypto 2018. See https://iacr.org/cryptodb/data/paper.php?pubkey=28802
1:07:51
Proving Theorems & Seeing Cats
Presented at SPLASH-I 2018
First off, you’re not going to learn anything about programmin...
published: 21 Nov 2018
Proving Theorems & Seeing Cats
Proving Theorems & Seeing Cats
- Report rights infringement
- published: 21 Nov 2018
- views: 100
Presented at SPLASH-I 2018
First off, you’re not going to learn anything about programming techniques, software development, or really, anything useful from this talk. I will tell you how a program I wrote for DARPA to help thwart terrorist plots turned into one that could write the following “haiku” when inspired by the phrase “loud guitar blues music”:
tuned adrenalin
my music
a beat-boogied headful
Along the way I will tell you two things:
how and why programming is stranger than you think
why real artificial intelligence needs to be a blend of proving theorems (in deference to John McCarthy) and seeing cats (in deference to Andrew Ng)
Geoffrey Jefferson, more or less debating Alan Turing, wrote: “Not until a machine can write a sonnet or compose a concerto because of thoughts and emotions felt, and not by the chance fall of symbols, could we agree that machine equals brain—that is, not only write it but know that it had written it.” [‘The Mind of Mechanical Man,’ British Medical Journal, 1949]
Stopping by Sums on a Snowy Mean Solar Day by the Roberts Frost and Harper
Whose sums these are I express I know.
His unit is in the body though;
He will not see me entering existing
To consider his sums operate up with snow.
My smaller horse must evaluate it interesting
To keep in without a family discriminating
Between the sums and polar form
The darkest mean solar day of the fundamental quantity.
He gives his influence devices an exit
To pass along if there is some statement.
The only other element is the use
Of benign context and yielding element.
The sums are logical, insensitive, and severe.
But I have expressions to position,
And large indefinite amounts to go before I sleep in,
And large indefinite amounts to go before I position.
1:22:59
Josef Urban | AI and Theorem Proving
1/13/2021 New Technologies in Mathematics Seminar
Speaker: Josef Urban, Czech Institute o...
published: 19 Jan 2021
Josef Urban | AI and Theorem Proving
Josef Urban | AI and Theorem Proving
- Report rights infringement
- published: 19 Jan 2021
- views: 3063
1/13/2021 New Technologies in Mathematics Seminar
Speaker: Josef Urban, Czech Institute of Informatics, Robotics and Cybernetics
Title: AI and Theorem Proving
Abstract: The talk will discuss the main approaches that combine machine learning with automated theorem proving and automated formalization. This includes learning to choose relevant facts for “hammer” systems, guiding the proof search of tableaux and superposition automated provers by interleaving learning and proving (reinforcement learning) over large ITP libraries, guiding the application of tactics in interactive tactical systems, and various forms of lemmatization and conjecturing. I will also show some demos of the systems, and discuss autoformalization approaches such as learning probabilistic grammars from aligned informal/formal corpora, combining them with semantic pruning, and using neural methods to learn direct translation from Latex to formal mathematics.
1:09:54
20190605 Introduction to Interactive theorem proving, at the OSU Quantum Symmetries summer school
published: 05 Jun 2019
20190605 Introduction to Interactive theorem proving, at the OSU Quantum Symmetries summer school
20190605 Introduction to Interactive theorem proving, at the OSU Quantum Symmetries summer school
- Report rights infringement
- published: 05 Jun 2019
- views: 1299
43:54
Kevin Buzzard: Mathematics and the Computer with G-Research
Kevin Buzzard, Professor of Pure Mathematics at Imperial College London, discusses Mathema...
published: 22 Dec 2022
Kevin Buzzard: Mathematics and the Computer with G-Research
Kevin Buzzard: Mathematics and the Computer with G-Research
- Report rights infringement
- published: 22 Dec 2022
- views: 3485
Kevin Buzzard, Professor of Pure Mathematics at Imperial College London, discusses Mathematics and the Computer as part of G-Research's Computer Guided Mathematics Symposium.
Held at the Science Museum in Central London, Kevin was the first of three speakers, alongside Alex Davies (DeepMind) and Sir Timothy Gowers (Professeur titulaire of the Combinatorics chair at the Collège de France).
Want to watch the talks? Check out our Computer Guided Mathematics Symposium playlist here: https://www.youtube.com/playlist?list=PL_9Cdw79RQrS65GRmrYaauOOygLpXdmMv
View Kevin's full presentation slides: http://bit.ly/3CHKYbc
The G-Research Computer Guided Mathematics Symposium was the final event in our 2022 Distinguished Speaker Series. We'll be running events in 2023, so if you're interested in attending future Distinguished Speaker Series events, please register your details here: https://events.beamery.com/gresearch/all-dss-events-mntauiaxr
5:04
Introducing PUTNAMBENCH: Revolutionizing AI Theorem Proving
UT Austin Researchers Launch PUTNAMBENCH: A Milestone in AI Benchmarking
Researchers at t...
published: 20 Jul 2024
Introducing PUTNAMBENCH: Revolutionizing AI Theorem Proving
Introducing PUTNAMBENCH: Revolutionizing AI Theorem Proving
- Report rights infringement
- published: 20 Jul 2024
- views: 37
UT Austin Researchers Launch PUTNAMBENCH: A Milestone in AI Benchmarking
Researchers at the University of Texas at Austin have unveiled PUTNAMBENCH, an innovative benchmark designed to evaluate neural theorem-provers by utilizing problems from the highly competitive William Lowell Putnam Mathematical Competition. This benchmark addresses a critical gap in AI development, where existing benchmarks focus too much on high-school-level mathematics and fail to challenge theorem-provers with advanced problems.
PUTNAMBENCH offers 1697 formalizations of 640 problems available in multiple formal proof languages, including Lean 4, Isabelle, and Coq. The meticulous creation of these problems aims to ensure rigorous testing across different theorem-proving environments. Despite current methods like GPT-4 and Sledgehammer being tested, the results indicate that only a few problems were successfully solved. This highlights the ongoing challenges faced in automating mathematical reasoning
With PUTNAMBENCH, researchers hope to drive future innovation and improvements in neural theorem provers. What are your thoughts on the impact of advanced benchmarks like PUTNAMBENCH on the future of AI in mathematical problem-solving?
Explore how Klap AI can transform your videos into viral short content at www.TheBestAI.org/now.
#AI #research #mathematics #innovation
Unveiling PUTNAMBENCH: The Ultimate AI Benchmark! by Steven's Workspace
OUTLINE:
00:00:00 Introducing PUTNAMBENCH
00:00:49 Automating Mathematical Reasoning
00:01:23 The Need for a Challenging Benchmark
00:02:37 The Genesis of PUTNAMBENCH
00:03:41 Multiple Formalisms in PUTNAMBENCH
00:04:51 The Potential of PUTNAMBENCH
Interactive Theorem Proving (conference)
Interactive Theorem Proving (ITP) is an annual international academic conference on the topic of automated theorem proving, proof assistants and related topics, ranging from theoretical foundations to implementation aspects and applications in program verification, security, and formalization of mathematics.
ITP brings together the communities using many systems based on higher-order logic such as ACL2, Coq, Mizar, HOL, Isabelle, NuPRL, PVS, and Twelf. Individual workshops or meetings devoted to individual systems are usually held concurrently with the conference.
Together with CADE and TABLEAUX, ITP is usually one of the three main conferences of the International Joint Conference on Automated Reasoning (IJCAR) whenever it convenes,
History
The inaugural meeting of ITP was held on 11–14 July 2010 in Edinburgh, Scotland, as part of the Federated Logic Conference. It is the extension of the Theorem Proving in Higher Order Logics (TPHOLs) conference series to the broad field of interactive theorem proving. TPHOLs meetings took place every year from 1988 until 2009.
This page contains text from Wikipedia, the Free Encyclopedia - https://wn.com/Interactive_Theorem_Proving_(conference)
Jason Rute - Deep learning in interactive theorem ...
Interactive Theorem Proving with Lean...
Towards Lean 4: An Optimized Object Model for an I...
CAIS-23-03 | Professor Lawrence Paulson | Automate...
Non Interactive Zero Knowledge Proofs for Composit...
Proving Theorems & Seeing Cats...
Josef Urban | AI and Theorem Proving...
20190605 Introduction to Interactive theorem provi...
Kevin Buzzard: Mathematics and the Computer with G...