Research: show 3 selected papers per theme, collapse the rest; update README contact

- Each theme shows its 3 most relevant papers from the strongest venues (the
  'representative' keys in _data/themes.yml) as full entries; the other lab-era papers
  sit under 'Show N more'. Replaces the separate representative cards (page ~35% shorter)
- Representative picks: UNCAP/CoRL'25/NeurIPS'23 (swarm), ICLR'25/MLSys'22/RSS'21 (edge),
  AAAI'26/CVPR'25/ECCV'24 (neuro-symbolic), NeurIPS'26/ICML'22/CoRL'22 (trust),
  NeurIPS'25/ICML'24/ICRA'23 (time series)
- README: help contact is now Oguzhan Baser (oguzhanbaser@utexas.edu); documents how to
  choose representative papers

Author: oguzhan-baser

README.md

@@ -251,12 +251,12 @@ The Research page is driven by two things: `_data/themes.yml` (the five themes,
   headline: Formal and cryptographic guarantees for deployed models.   # one line, also on Home
   blurb: >-                        # the paragraph under the title
     Deployed models face adversaries, ...
-  representative: [baser2026tensorcommitments, agarwal2022synthesizing, baser2025phonemefake]   # bib keys, 3 shown as cards
+  representative: [baser2026tensorcommitments, cheng2022task, agarwal2022synthesizing]   # 3 bib keys shown as "Selected papers"; the rest are collapsed
 ```
 
 - Papers from **2021 on** are listed under the **primary** (first) theme.
 - Papers before 2021 go into the "Before Swarm Lab" block, grouped by theme; `themes={other}` goes into "Earlier & other work".
-- `representative` takes bib keys that must exist in `references.bib`.
+- `representative` takes bib keys that must exist in `references.bib`. Pick the most relevant papers from the strongest venues, from 2021 on, whose **primary** theme is this theme (otherwise the "Show N more" count is off).
 - Slugs must match exactly. Adding a sixth theme is a design decision: also check `research.md`, where each theme has its own small illustration, and ask first.
 
 ## Write a blog post
@@ -323,7 +323,7 @@ Never push straight to a branch other than `source`; there is no separate `maste
 
 **Something broke?** Open the failed run in the Actions tab and read the "Build with Jekyll" log. Typical causes: a duplicate BibTeX key, an unbalanced `{` in a bib entry, bad YAML in `news.yml` or a front matter block (check indentation and quotes), or a post without a `YYYY-MM-DD-` prefix.
 
-**Help:** text Po-han in the Swarm Lab Slack or email pohanli@utexas.edu; if that does not work, ping Sundar.
+**Help:** message Oguzhan Baser in the Swarm Lab Slack or email oguzhanbaser@utexas.edu.
 
 ## Credits
 

_data/themes.yml

@@ -8,7 +8,7 @@
     We develop game-theoretic, submodular, and multi-agent learning methods that decide how robots
     plan together, which data each robot collects and shares, and where human supervision goes,
     so the whole swarm improves as one.
-  representative: [akcin2025dual, bhatt2025uncap, akcin2023fleet]
+  representative: [bhatt2025uncap, akcin2025dual, li2023taskaware]
 - slug: edge
   name: Edge–Cloud Systems for Embodied AI
   headline: Communication-efficient perception and inference across networks.
@@ -17,7 +17,7 @@
     learning with communication through task-aware compression, offloading, adaptive model
     selection, and resource allocation, together with representations that make robot perception
     efficient at the edge.
-  representative: [li2023taskaware, baser2025fairsynergy, Omama2024exploiting]
+  representative: [Omama2024exploiting, qiu2022ml, nakanoya2021co]
 - slug: neurosymbolic
   name: Neuro-Symbolic Video Understanding
   headline: Grounding foundation models in temporal logic.
@@ -33,12 +33,12 @@
     Deployed models face adversaries, privacy leakage, deepfakes, and untrusted compute providers.
     We build verifiable LLM inference, privacy-aware representations, deepfake defenses, and robust
     controllers under adversarial conditions.
-  representative: [baser2026tensorcommitments, agarwal2022synthesizing, baser2025phonemefake]
+  representative: [baser2026tensorcommitments, cheng2022task, agarwal2022synthesizing]
 - slug: timeseries
   name: Time-Series Generation & Forecasting
   headline: Conditional, constrained generation of temporal data.
   blurb: >-
     Physical systems are described by time series. We build conditional and constrained
     time-series generators and forecasts that are robust, private, and evaluated by the decisions
     they support.
-  representative: [narasimhan2024time, agarwal2023, li2022differentially]
+  representative: [narasimhan2024time, narasimhan2024cps, agarwal2023]

_sass/_research.scss

@@ -195,58 +195,6 @@
   white-space: nowrap;
 }
 
-// ---------- Representative papers ----------
-.reps {
-  display: grid;
-  grid-template-columns: repeat(auto-fit, minmax(min(100%, 12.5rem), 1fr));
-  gap: var(--space-s);
-
-  > ul, li { margin: 0; padding: 0; list-style: none; min-width: 0; block-size: 100%; }
-  .pub-list > li { padding: 0; border: 0; }
-}
-
-.rep {
-  display: flex;
-  flex-direction: column;
-  gap: var(--space-2xs);
-  block-size: 100%;
-  padding: var(--space-s);
-  border: var(--border-w) solid var(--border);
-  border-radius: var(--radius-m);
-  background: var(--surface);
-  color: var(--text);
-  text-decoration: none;
-  transition: border-color var(--dur-fast) var(--ease-out), box-shadow var(--dur-fast) var(--ease-out),
-    transform var(--dur-fast) var(--ease-out);
-
-  &:hover { border-color: var(--accent); box-shadow: var(--shadow-2); transform: translateY(-2px); text-decoration: none; }
-  &:focus-visible { outline: var(--focus); outline-offset: var(--focus-offset); }
-}
-
-.rep__meta {
-  display: flex;
-  flex-wrap: wrap;
-  gap: var(--space-3xs) var(--space-2xs);
-  font-family: var(--font-mono);
-  font-size: var(--step--1);
-}
-
-.rep__venue { color: var(--accent-text); font-weight: var(--weight-medium); }
-
-.rep__award {
-  padding: 0 0.5rem;
-  border-radius: var(--radius-pill);
-  background: var(--accent-tint);
-  color: var(--accent-strong);
-  &:empty { display: none; }
-}
-
-.rep__title {
-  font-weight: var(--weight-semibold);
-  line-height: var(--lh-snug);
-  text-wrap: balance;
-}
-
 // ---------- Full lists ----------
 .theme-section {
   .pub-title { font-size: var(--step-0); }
@@ -330,6 +278,5 @@
 }
 
 @media (prefers-reduced-motion: reduce) {
-  .rep, .research-index__link, .more__summary, .more__summary::before { transition: none; }
-  .rep:hover { transform: none; }
+  .research-index__link, .more__summary, .more__summary::before { transition: none; }
 }

research.md

@@ -7,7 +7,7 @@ lede: We build learning, reasoning, and verification for fleets of robots and AI
 {%- comment -%}
   Data: _data/themes.yml (order, text, representative keys) + `themes` field in the bib (first slug = primary).
   Lab era = year >= 2021, "Before Swarm Lab" = year < 2021, `other` = Earlier & other work.
-  Entry markup comes from _layouts/bib.html; the compact "representative" cards use an inline template (-T).
+  Entry markup comes from _layouts/bib.html. `representative` keys must be lab-era papers whose primary theme is this theme.
 {%- endcomment -%}
 <div class="research">
   <nav class="research-index" aria-label="Research themes">
@@ -58,22 +58,20 @@ lede: We build learning, reasoning, and verification for fleets of robots and AI
         <p class="theme-section__blurb">{{ t.blurb }}</p>
       </header>
 
-      {%- if t.representative %}
-      <h3 class="theme-section__label">Representative papers</h3>
-      <div class="reps">
-        {%- for k in t.representative %}
-        {% bibliography --group_by none -T <a class="rep" href="{{ entry.url | default: entry.arxiv | default: entry.pdf | default: entry.project }}"><span class="rep__meta"><span class="rep__venue">{{ entry.venue_short }}</span><span class="rep__award">{{ entry.award }}</span></span><span class="rep__title">{{ entry.title }}</span></a> --query @*[key={{ k }}] %}
-        {%- endfor %}
-      </div>
-      {%- endif %}
-
-      <h3 class="theme-section__label">All papers since 2021 <span class="theme-section__count">{{ n }} {% if n == 1 %}paper{% else %}papers{% endif %}</span></h3>
-      {% bibliography --group_by none --max 6 --query @*[themes^={{ t.slug }}, year>=2021] %}
-      {%- if n > 6 %}
-      {%- assign rest = n | minus: 6 %}
+      {%- comment -%} Selected = the theme's `representative` keys (most relevant, top venues); the rest of the lab-era papers are collapsed. {%- endcomment -%}
+      {%- assign sel = "" %}{%- assign excl = "" %}
+      {%- for k in t.representative %}
+        {%- unless forloop.first %}{% assign sel = sel | append: " || " %}{% endunless %}
+        {%- assign sel = sel | append: "key=" | append: k %}
+        {%- assign excl = excl | append: ", key!=" | append: k %}
+      {%- endfor %}
+      <h3 class="theme-section__label">Selected papers <span class="theme-section__count">{{ n }} {% if n == 1 %}paper{% else %}papers{% endif %} since 2021</span></h3>
+      {% bibliography --group_by none --query @*[{{ sel }}] %}
+      {%- assign rest = n | minus: t.representative.size %}
+      {%- if rest > 0 %}
       <details class="more">
         <summary class="more__summary"><span class="more__closed">Show {{ rest }} more</span><span class="more__open">Show fewer</span></summary>
-        {% bibliography --group_by none --offset 6 --query @*[themes^={{ t.slug }}, year>=2021] %}
+        {% bibliography --group_by none --query @*[themes^={{ t.slug }}, year>=2021{{ excl }}] %}
       </details>
       {%- endif %}
     </section>