Merge branch 'compiler-research:master' into master

Author: GalinBistrev2

.github/actions/spelling/allow/terms.txt

@@ -1,5 +1,6 @@
 AARCH
 AIML
+Backpropagation
 BGZF
 CINT
 CMSSW
@@ -16,6 +17,7 @@ IIT
 JIT'd
 Jacobians
 Jurgaityt
+LHC
 LLMs
 LLVM
 LULESH
@@ -49,6 +51,7 @@ jthread
 linkedin
 llm
 llvm
+meetinglist
 microenvironments
 omp
 openmp

_data/contributors.yml

@@ -159,8 +159,16 @@
   github: "https://github.com/ovdiiuv"
   active: 1
   projects:
-    - title: "Optimizing automatic differentiation using activity analysis"
+    - title: "Activity analysis for reverse-mode differentiation of (CUDA) GPU kernels"
       status: Ongoing
+      description: |
+        Clad is a Clang plugin designed to provide automatic differentiation (AD) for C++ mathematical functions. It generates code for computing derivatives modifying Abstract-Syntax-Tree(AST) using LLVM compiler features. It performs advanced program optimization by implementing more sophisticated analyses because it has access to a rich program representation – the Clang AST. Clad supports reverse-mode differentiation of the CUDA kernels, however it is not always optimal because the generated code might contain the data-race conditions, significantly slowing up the execution. Thread Safety Analysis is a static analysis that detects possible data-race conditions that would enable reducing atomic operations in the Clad-produced code.
+
+      proposal: /assets/docs/Maksym_Andriichuk_Proposal_2025.pdf
+      mentors: Vassil Vassilev, David Lange
+
+    - title: "Optimizing automatic differentiation using activity analysis"
+      status: Completed project
       description: |
         Clad has undergone significant optimization in the recent years. One such optimization is the To-Be-Recorded optimization which reduces the memory pressure to the clad tape data structure. The next step is to implement the activity analysis which discards all statements which are irrelevant for the generated code. That is, if the statements do not depend on the input/output variables of a routine in a differentiable way, they are ignored. Combining both analyses could result in a much greater performance, which is our ultimate goal.        
       proposal: /assets/docs/Maksym_Andriichuk_Proposal_2024.pdf

_data/meetinglist.yml

@@ -1,3 +1,39 @@
+- name: "CaaS Monthly Meeting"
+  date: 2025-06-05 17:00:00 +0200
+  time_cest: "17:00"
+  connect: "[Link to zoom](https://princeton.zoom.us/j/94431046845?pwd=D5i77Qb0PgfwwIubvbo2viEunne7eQ.1)"
+  label: caas_05June2025
+  agenda:
+    - title: "Differentiating RooFit likelihoods with Clad"
+      speaker:
+        image: "https://compiler-research.org/images/team/JRembser.png"
+        name: "Jonas Rembser"
+        bio: |
+          Jonas Rembser is a particle physicist with a strong interest in data
+          science and software development. After working with the CMS
+          experiment for his PhD in particle physics, Jonas joined CERN in 2021
+          to work on the ROOT project. He is takes care of the statistical
+          analysis libraries in ROOT, such as RooFit and Minuit 2, as well as
+          the ROOT Python interface. His other academic interests include
+          differential programming and financial markets.
+      description: |
+        RooFit is a C++ software for statistical data modeling that is part of
+        ROOT. It is widely used by the High Energy Physics (HEP) community,
+        most notably the LHC experiments. RooFits core feature is building and
+        numerically minimizing likelihood functions for parameter estimation
+        (aka. *fitting*). The minimization can be greatly sped up by using
+        reverse mode (Automatic Differentiation) AD, in particular for fits
+        with many parameters, as typical for the LHC experiments. This
+        presentation will tell the story of how AD was integrated to RooFit
+        powered by [Clad](https://github.com/vgvassilev/clad/), the AD tool
+        developed by Compiler Research. We will cover the integration strategy,
+        explain the particularities of AD in the context of likelihood fits,
+        and present benchmark results. The presentation will wrap up by
+        explaining the current limitations and plans for the future. More
+        information on AD in RooFit with Clad can also be found in
+        [this blog post](https://root.cern/blog/roofit-ad/) on the ROOT
+        website.
+      slides: /assets/presentations/J_Rembser-RooFit_Clad.pdf
 - name: "CaaS Monthly Meeting"
   date: 2025-05-15 17:00:00 +0200
   time_cest: "17:00"

_data/standing_meetings.yml

@@ -3,6 +3,14 @@
   time_cest: "17:00"
   connect: "[Link to zoom](https://princeton.zoom.us/j/97915651167?pwd=MXJ1T2lhc3Z5QWlYbUFnMTZYQlNRdz09)"
   agenda:
+    - title: "Enhancing LLM Training Efficiency with Clad for Backpropagation"
+      date: 2025-06-05 15:00:00 +0200
+      speaker: "Rohan Timmaraju"
+      link: "[Slides](/assets/presentations/CaaS_Weekly_05_06_2025_Rohan_Timmaraju_LLM_Training.pdf)"
+    - title: "Improve automatic differentiation of object-oriented paradigms using Clad"
+      date: 2025-06-05 16:00:00 +0200
+      speaker: "Petro Zarytskyi"
+      link: "[Slides](/assets/presentations/CaaS_Weekly_05_06_2025_Petro_Zarytskyi_AD_Object_Oriented_Paradigms.pdf)"
     - title: "Using ROOT in the field of genome sequencing"
       date: 2025-05-21 17:00:00 +0200
       speaker: "Aditya Pandey"

_posts/2025-05-21-enhancing-llm-training.md

@@ -43,7 +43,7 @@ By successfully integrating Clad into a C++ LLM training pipeline, we aim to:
 * **Offer a C++ Alternative:** Provide a foundation for more efficient, compiler-driven LLM training within the C++ ecosystems.
 * **Learn and Share:** Gain insights into the practicalities of applying compiler-based AD to complex ML problems and share these learnings with the community.
 
-I believe this project has the potential to make a valuable contribution to both the compiler research field and the ongoing efforts to make powerful AI models more accessible and efficient to train.
+This project has the potential to make a valuable contribution to both the compiler research field and the ongoing efforts to make powerful AI models more accessible and efficient to train.
 
 ### Related Links