Update dickens-arxiv24 to dickens-icml24.

_data/pubs/dickens-arxiv24.json → _data/pubs/dickens-icml24.json

@@ -1,5 +1,5 @@
 {
-  "type": "unpublished",
+  "type": "conference",
   "title": "Convex and Bilevel Optimization for Neuro-Symbolic Inference and Learning",
   "authors": [
     "Charles Dickens",
@@ -8,12 +8,14 @@
     "Stephen Wright",
     "Lise Getoor"
   ],
-  "venue": "arXiv",
+  "venue": "International Conference on Machine Learning",
   "year": "2024",
+  "publisher": "PMLR",
+  "address": "Vienna Austria",
   "links": [
     {
       "label": "paper",
-      "href": "/assets/resources/dickens-arxiv24.pdf"
+      "href": "/assets/resources/dickens-icml24.pdf"
     }
   ],
   "abstract": "We address a key challenge for neuro-symbolic (NeSy) systems by leveraging convex and bilevel optimization techniques to develop a general gradient-based framework for end-to-end neural and symbolic parameter learning. The applicability of our framework is demonstrated with NeuPSL, a state-of-the-art NeSy architecture. To achieve this, we propose a smooth primal and dual formulation of NeuPSL inference and show learning gradients are functions of the optimal dual variables. Additionally, we develop a dual block coordinate descent algorithm for the new formulation that naturally exploits warm-starts. This leads to over 100× learning runtime improvements over the current best NeuPSL inference method. Finally, we provide extensive empirical evaluations across 8 datasets covering a range of tasks and demonstrate our learning framework achieves up to a 16% point prediction performance improvement over alternative learning methods.",