Spring updates

Author: arvind

src/lib/data/news.json

@@ -1,4 +1,12 @@
 [
+  {
+    "date": "2026-05-14",
+    "desc": "Huge congratulations to [Dr. Angie Boggust](https://angieboggust.com) for a [masterful defense](https://x.com/mitvis/status/2054993532549468562) of her thesis on \"Reconciling AI Representations and User Mental Models\""
+  },
+  {
+    "date": "2026-04-22",
+    "desc": "Two papers this spring: Angie's [Semantic Regexes](/pubs/semantic-regexes) will appear at ICLR, and Matt's [Belidor](/pubs/belidor/) will appear at CHI."
+  },
   {
     "date": "2025-08-15",
     "desc": "It's going to be a jam-packed Fall with [Benthic](/pubs/benthic/) at ASSETS, [FabObscura](/pubs/fabobscura/) at UIST, and [GoFish](/pubs/gofish/), [Visualization Vibes](/pubs/vibes/), and (honorable-mention-winning!) [Quantifying Visualization Vibes](/pubs/quantifying-vibes/) at VIS!!"

src/lib/data/people.json

@@ -80,9 +80,9 @@
     "url": "https://ryanyen2.github.io/",
     "title": "PhD Researcher"
   },
-  "ticha": {
-    "name": "Ticha Sethapakdi",
-    "url": "https://www.tichamelody.com/",
+  "helenagv": {
+    "name": "Helena Vasconcelos",
+    "url": "https://helenavasc.com/",
     "title": "PhD Researcher"
   },
   "jonsh": {
@@ -259,20 +259,17 @@
     "affiliation": "IIT Kharagpur",
     "alumni": true
   },
-  "helenagv": {
-    "name": "Helena Vasconcelos",
-    "url": "https://helenavasc.com/",
-    "title": "Research Intern",
-    "affiliation": "Stanford University",
-    "alumni": true,
-    "next": "PhD Student at Harvard CS"
-  },
   "peck": {
     "name": "Evan Peck",
     "title": "Visiting Scientist",
     "url": "http://www.eg.bucknell.edu/~emp017",
     "alumni": true
   },
+  "ticha": {
+    "name": "Ticha Sethapakdi",
+    "url": "https://www.tichamelody.com/",
+    "title": "Visiting PhD Researcher"
+  },
   "berger": {
     "name": "Caroline Berger",
     "title": "Visiting PhD Researcher",
@@ -299,7 +296,8 @@
   "zdesimone": {
     "name": "Zoe De Simone",
     "url": "https://zoedesimone.github.io//",
-    "title": "PhD Researcher"
+    "title": "PhD Researcher",
+    "alumni": true
   },
   "arnavv": {
     "name": "Arnav Verma",

src/lib/data/venues.json

@@ -246,5 +246,13 @@
     },
     "html": "2023-03-01",
     "paywall": false
+  },
+  "iclr": {
+    "short": "ICLR",
+    "full": "International Conference on Learning Representations",
+    "bibtex": {
+      "type": "inproceedings",
+      "venue": "booktitle"
+    }
   }
 }

src/routes/pubs/belidor.md

@@ -0,0 +1,31 @@
+---
+title: 'Belidor: A Specification Language for Operationalizing Structural Analogies Between User Interfaces'
+authors:
+  - key: mattbl
+    affiliation: UC San Diego
+  - name: Devamardeep Hayatpur
+    affiliation: UC San Diego
+  - key: arvindsatya
+  - name: Haijun Xia
+    affiliation: UC San Diego
+venue: chi
+date: 2026-04-13
+doi: 10.1145/3772318.3791613
+themes: 
+  - convivial-interaction
+tags:
+  - language
+  - qualitative methods
+feature: true
+teaser: >
+  Belidor specifications describe the structure underlying user interfaces. Events in a calendar and videos in a video editor are both organized by their respective temporal structures (1). Conceptual objects and structures can be presented by mapping them to the view’s objects (marks) and structures (encodings) (2). Structures also characterize interaction: for example, events/videos are directly manipulated along the time/timeline structure (3). Belidor specifications help compare and contrast user interfaces, in this case by aligning the parts of the calendar/video editor specifications that overlap (4), highlighting the analogy between the two user interfaces.
+materials:
+  - name: Analogy Viewer
+    url: https://belidor-specification.github.io/
+    type: laptop
+---
+We present Belidor, a text notation that describes the structure underlying user interfaces (UIs). Belidor’s relational model emphasizes how structures, such as the temporal order of text messages, cut across an interactive system’s conceptual model, user-facing presentation, and interactive behavior. We demonstrate Belidor’s expressive power with a gallery of examples spanning GUIs (eg. messaging, video editors), screen readers, and hardware devices.
+
+Belidor serves as an effective representation for structural analogies between user interfaces (eg. between calendars and video-editors). In contrast, prior work relied on visual UI representations and therefore prioritized visual style transfer.
+In three case studies, we show how Belidor can reveal analogies, help transfer ideas between user interfaces, and describe design patterns as analogies
+We discuss the implications of representing the structure of interactive systems for designers and developers, and envision how Belidor might support “structural design moves” for interface designers.

src/routes/pubs/semantic-regexes.md

@@ -0,0 +1,34 @@
+---
+title: 'Semantic Regexes: Auto-Interpreting LLM Features with a Structured Language'
+authors:
+  - key: aboggust
+  - name: Donghao Ren
+    affiliation: Apple
+  - name: Yannick Assogba
+    affiliation: Apple
+  - name: Dominik Moritz
+    affiliation: Apple
+  - key: arvindsatya
+  - name: Fred Hohman
+    affiliation: Apple
+venue: iclr
+date: 2026-04-24
+doi: 10.48550/arXiv.2510.06378
+themes: 
+  - interpretability
+tags:
+  - language
+feature: true
+teaser: Semantic regexes are often more concise (top), more consistently describe equivalent features (middle), and better reflect feature complexity (bottom) than natural language descriptions.
+materials:
+  - name: Viewer
+    url: https://apple.github.io/ml-semantic-regex/
+    type: chart-line
+  - name: Python Package
+    url: https://pypi.org/project/semantic-regex/
+    type: brands fa-python
+  - name: Code Repo
+    url: https://github.com/apple/ml-semantic-regex
+    type: code
+---
+Automated interpretability aims to translate large language model (LLM) features into human understandable descriptions. However, natural language feature descriptions can be vague, inconsistent, and require manual relabeling. In response, we introduce semantic regexes, structured language descriptions of LLM features. By combining primitives that capture linguistic and semantic patterns with modifiers for contextualization, composition, and quantification, semantic regexes produce precise and expressive feature descriptions. Across quantitative benchmarks and qualitative analyses, semantic regexes match the accuracy of natural language while yielding more concise and consistent feature descriptions. Their inherent structure affords new types of analyses, including quantifying feature complexity across layers, scaling automated interpretability from insights into individual features to model-wide patterns. Finally, in user studies, we find that semantic regexes help people build accurate mental models of LLM features.