Patrick Shafto
@patrickshafto.bsky.social
Prof: Rutgers; Program Manager: DARPA; Scientist-at-Large: Redpoll.
Math! Machine learning! Cognitive Science!
Math! Machine learning! Cognitive Science!
Exciting opportunity for people at the intersection of math, AI, and formal methods!
University of Michigan Math search: advanced assistant, or tenured associate or full professor for our initiative on Foundations of Artificial Intelligence
mathjobs.org/jobs/list/26...
University of Michigan Math search: advanced assistant, or tenured associate or full professor for our initiative on Foundations of Artificial Intelligence
mathjobs.org/jobs/list/26...
October 8, 2025 at 1:40 PM
Exciting opportunity for people at the intersection of math, AI, and formal methods!
University of Michigan Math search: advanced assistant, or tenured associate or full professor for our initiative on Foundations of Artificial Intelligence
mathjobs.org/jobs/list/26...
University of Michigan Math search: advanced assistant, or tenured associate or full professor for our initiative on Foundations of Artificial Intelligence
mathjobs.org/jobs/list/26...
AI and math. Geometry and symbolic reasoning. Amazing recent developments and stellar line up of speakers. It is going to be an exciting week!
The Geometry of Machine Learning @ Harvard Center for Mathematical Sciences and Applications (CMSA)
cmsa.fas.harvard.edu/event/mlgeom...
The Geometry of Machine Learning @ Harvard Center for Mathematical Sciences and Applications (CMSA)
cmsa.fas.harvard.edu/event/mlgeom...
September 15, 2025 at 12:49 AM
AI and math. Geometry and symbolic reasoning. Amazing recent developments and stellar line up of speakers. It is going to be an exciting week!
The Geometry of Machine Learning @ Harvard Center for Mathematical Sciences and Applications (CMSA)
cmsa.fas.harvard.edu/event/mlgeom...
The Geometry of Machine Learning @ Harvard Center for Mathematical Sciences and Applications (CMSA)
cmsa.fas.harvard.edu/event/mlgeom...
Spent last week @ias with DeepMind math teams. AlphaProof, AlphaGeometry, PDEs, and PAIRs teams were there with leadership. Impressed by the collaboration. Huge, open technical problems to be solved. Exciting times for mathematics & AI!
www.ias.edu/math/events/...
www.ias.edu/math/events/...
May 6, 2025 at 12:03 PM
Spent last week @ias with DeepMind math teams. AlphaProof, AlphaGeometry, PDEs, and PAIRs teams were there with leadership. Impressed by the collaboration. Huge, open technical problems to be solved. Exciting times for mathematics & AI!
www.ias.edu/math/events/...
www.ias.edu/math/events/...
The Exponentiating Mathematics (expMath) program BAA has dropped!
sam.gov/opp/869c8d73...
Abstracts due May 15. Be sure to register on the submission website well in advance! (A few days ahead!)
sam.gov/opp/869c8d73...
Abstracts due May 15. Be sure to register on the submission website well in advance! (A few days ahead!)
April 30, 2025 at 6:49 PM
The Exponentiating Mathematics (expMath) program BAA has dropped!
sam.gov/opp/869c8d73...
Abstracts due May 15. Be sure to register on the submission website well in advance! (A few days ahead!)
sam.gov/opp/869c8d73...
Abstracts due May 15. Be sure to register on the submission website well in advance! (A few days ahead!)
University senate
April 19, 2025 at 4:04 PM
University senate
Proposer's day for expMath has dropped!
Accelerate progress in pure math via AI capable of proposing and proving useful abstractions
Teams either:
- develop AI capable of auto decomposition and auto(in)formalization
- evaluate with respect to professional math
Accelerate progress in pure math via AI capable of proposing and proving useful abstractions
Teams either:
- develop AI capable of auto decomposition and auto(in)formalization
- evaluate with respect to professional math
March 26, 2025 at 8:24 PM
Proposer's day for expMath has dropped!
Accelerate progress in pure math via AI capable of proposing and proving useful abstractions
Teams either:
- develop AI capable of auto decomposition and auto(in)formalization
- evaluate with respect to professional math
Accelerate progress in pure math via AI capable of proposing and proving useful abstractions
Teams either:
- develop AI capable of auto decomposition and auto(in)formalization
- evaluate with respect to professional math
DARPA released a Request for Information (RFI) that seeks community feedback on the draft DARPA Guide to Formal Methods to Deliver Resilient Systems for Proposals (“the FMDRS Guide”). You can find the RFI here on Sam.gov.
Details in the image...
Details in the image...
February 14, 2025 at 10:04 PM
DARPA released a Request for Information (RFI) that seeks community feedback on the draft DARPA Guide to Formal Methods to Deliver Resilient Systems for Proposals (“the FMDRS Guide”). You can find the RFI here on Sam.gov.
Details in the image...
Details in the image...
Looking forward to this talk by Leonardo de Moura!
@leanprover
www.simonsfoundation.org/event/verifi...
@leanprover
www.simonsfoundation.org/event/verifi...
February 14, 2025 at 2:33 PM
Looking forward to this talk by Leonardo de Moura!
@leanprover
www.simonsfoundation.org/event/verifi...
@leanprover
www.simonsfoundation.org/event/verifi...
Simons Institute special year on Language models and transformers, includes some great upcoming workshops:
The future of LLMs
AI for math
Safety guarantees
Theory of trustworthy AI
So stoked that we @ darpa are supporting this amazing work.
simons.berkeley.edu/programs/spe...
The future of LLMs
AI for math
Safety guarantees
Theory of trustworthy AI
So stoked that we @ darpa are supporting this amazing work.
simons.berkeley.edu/programs/spe...
February 12, 2025 at 4:32 PM
Simons Institute special year on Language models and transformers, includes some great upcoming workshops:
The future of LLMs
AI for math
Safety guarantees
Theory of trustworthy AI
So stoked that we @ darpa are supporting this amazing work.
simons.berkeley.edu/programs/spe...
The future of LLMs
AI for math
Safety guarantees
Theory of trustworthy AI
So stoked that we @ darpa are supporting this amazing work.
simons.berkeley.edu/programs/spe...
Super excited: my new @darpa program on AI for pure mathematics!
Exponentiating Mathematics (expMath) aims to accelerate the rate of progress in pure math through the development of an AI collaborator and new professional-level math benchmarks.
sam.gov/opp/4def3c13...
Exponentiating Mathematics (expMath) aims to accelerate the rate of progress in pure math through the development of an AI collaborator and new professional-level math benchmarks.
sam.gov/opp/4def3c13...
February 7, 2025 at 4:58 PM
Super excited: my new @darpa program on AI for pure mathematics!
Exponentiating Mathematics (expMath) aims to accelerate the rate of progress in pure math through the development of an AI collaborator and new professional-level math benchmarks.
sam.gov/opp/4def3c13...
Exponentiating Mathematics (expMath) aims to accelerate the rate of progress in pure math through the development of an AI collaborator and new professional-level math benchmarks.
sam.gov/opp/4def3c13...
Now is your chance!
"Action gaps and advantages in continuous time distributional RL"
Poster 6410 #NeurIPS2024
"Action gaps and advantages in continuous time distributional RL"
Poster 6410 #NeurIPS2024
December 13, 2024 at 7:28 PM
Now is your chance!
"Action gaps and advantages in continuous time distributional RL"
Poster 6410 #NeurIPS2024
"Action gaps and advantages in continuous time distributional RL"
Poster 6410 #NeurIPS2024
Apply to participate in our "Mathematics of intelligences" at
the IPAM at UCLA Sept 9-Dec 13 2024!
One can also apply to just "Workshop II: Theory and Practice of Deep Learning" Oct 14-18 (or the other workshops!) Details and application: www.ipam.ucla.edu/programs/lon...
the IPAM at UCLA Sept 9-Dec 13 2024!
One can also apply to just "Workshop II: Theory and Practice of Deep Learning" Oct 14-18 (or the other workshops!) Details and application: www.ipam.ucla.edu/programs/lon...
February 1, 2024 at 4:09 PM
Apply to participate in our "Mathematics of intelligences" at
the IPAM at UCLA Sept 9-Dec 13 2024!
One can also apply to just "Workshop II: Theory and Practice of Deep Learning" Oct 14-18 (or the other workshops!) Details and application: www.ipam.ucla.edu/programs/lon...
the IPAM at UCLA Sept 9-Dec 13 2024!
One can also apply to just "Workshop II: Theory and Practice of Deep Learning" Oct 14-18 (or the other workshops!) Details and application: www.ipam.ucla.edu/programs/lon...
The parameter space, visualized as a cube, with special points that correspond to existing machine learning models labeled.
October 27, 2023 at 4:27 PM
The parameter space, visualized as a cube, with special points that correspond to existing machine learning models labeled.
We visualize the space of learning models encoded by GBT as a cube which includes classic learning models as special points. We derive critical properties of this parameterized space including proving continuity and differentiability which is the basis for model interpolation.
October 27, 2023 at 4:26 PM
We visualize the space of learning models encoded by GBT as a cube which includes classic learning models as special points. We derive critical properties of this parameterized space including proving continuity and differentiability which is the basis for model interpolation.
*NEW* "Generalized Belief Transport" Led by Junqi Wang and Pei Wang, accepted to
@NeurIPSConf
#NeurIPS2023! In ML we often assume that is fixed and known, but agents need to be able to change with constraints on knowledge, time, drift, ...
shaftolab.com/assets/paper...
@NeurIPSConf
#NeurIPS2023! In ML we often assume that is fixed and known, but agents need to be able to change with constraints on knowledge, time, drift, ...
shaftolab.com/assets/paper...
October 27, 2023 at 4:25 PM
*NEW* "Generalized Belief Transport" Led by Junqi Wang and Pei Wang, accepted to
@NeurIPSConf
#NeurIPS2023! In ML we often assume that is fixed and known, but agents need to be able to change with constraints on knowledge, time, drift, ...
shaftolab.com/assets/paper...
@NeurIPSConf
#NeurIPS2023! In ML we often assume that is fixed and known, but agents need to be able to change with constraints on knowledge, time, drift, ...
shaftolab.com/assets/paper...
We show that our approach is generalizes a vast majority of previous models! (i.e. they are special cases where common ground is perfect.)
October 22, 2023 at 1:12 AM
We show that our approach is generalizes a vast majority of previous models! (i.e. they are special cases where common ground is perfect.)
We define common ground which is not perfect, and cooperative communication in such settings.
October 22, 2023 at 1:05 AM
We define common ground which is not perfect, and cooperative communication in such settings.
*NEW* "Common ground in cooperative communication" led by @XiaoranHao and Yash Jhaveri has been accepted to @NeurIPSConf! #NeurIPS2023
We introduce the first mathematical definition of common ground, show it generalizes previous models, and derive a connection to the new cVAE!
We introduce the first mathematical definition of common ground, show it generalizes previous models, and derive a connection to the new cVAE!
October 22, 2023 at 1:02 AM
*NEW* "Common ground in cooperative communication" led by @XiaoranHao and Yash Jhaveri has been accepted to @NeurIPSConf! #NeurIPS2023
We introduce the first mathematical definition of common ground, show it generalizes previous models, and derive a connection to the new cVAE!
We introduce the first mathematical definition of common ground, show it generalizes previous models, and derive a connection to the new cVAE!