Skip to content

Commit 20a7300

Browse files
Guanya ShiGuanya Shi
Guanya Shi
authored and
Guanya Shi
committed
iclr-l4dc
1 parent b344c53 commit 20a7300

2 files changed

Lines changed: 93 additions & 71 deletions

File tree

‎index.html‎

Lines changed: 6 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -109,12 +109,18 @@
109109
<tr>
110110
<td width="100%" valign="middle">
111111
<ul>
112+
<li>2026-01-26: Four papers accepted to ICLR 2026 and two to L4DC 2026.</li>
112113
<li>2025-10-21: Congratulations to Zeji for receiving the <a href="https://www.cs.cmu.edu/news/2025/amazon-phd-fellows" target="_blank">Amazon AI Ph.D. Fellowship</a>, and Wenli for receiving the CMU RI Presidential Fellowship!</li>
113114
<li>2025-08-01: <a href="https://lecar-lab.github.io/spi-active_/" target="_blank">SPI-Active</a> (oral), <a href="https://lecar-lab.github.io/SoFTA/" target="_blank">HoldMyBeer</a>, and <a href="https://human-as-robot.github.io/" target="_blank">Humanoid~Human</a> accepted to CoRL 2025.</li>
114115
<li>2025-04-11: <a href="https://agile.human2humanoid.com/" target="_blank">ASAP</a> and <a href="https://xiaofeng-guo.github.io/flying_hand.io/" target="_blank">Flying Hand</a> accepted to RSS 2025.</li>
115116
<li>2025-02-10: Congratulations to Zeji for receiving the CMU Wei Shen and Xuehong Zhang Presidential Fellowship, and Chaoyi for receiving the Hsu Chang Memorial Fellowship!</li>
116117
<li>2025-01-30: 8 papers accepted to ICRA 2025 and <a href="https://lecar-lab.github.io/dial-mpc/" target="_blank">DIAL-MPC</a> is a Best Paper Award Finalist!</li>
117118
<li>2024-12-18: Congratulations to Tairan for receiving the CMU RI Presidential Fellowship and the <a href="https://blogs.nvidia.com/blog/graduate-fellowship-recipients-2025-2026/" target="_blank">NVIDIA Fellowship</a>!</li>
119+
</ul>
120+
121+
<details>
122+
<summary>See More</summary></li>
123+
<ul>
118124
<li>2024-10-15: Guanya gave talks at <a href="https://www.ri.cmu.edu/event/building-generalist-robots-with-agility-via-learning-and-control-humanoids-and-beyond/" target="_blank">CMU RI Seminar</a> and <a href="https://ai.engin.umich.edu/events/2024-ai-symposium/" target="_blank">Michigan AI Symposium</a>. Check out the <a href="https://youtu.be/Uym3Tr6t5TM" target="_blank">recording</a>.</li>
119125
<li>2024-10-09: <a href="https://mediaspace.gatech.edu/media/1_tckra9zz" target="_blank">Flying Calligrapher</a> accepted to RA-L.</li>
120126
<li>2024-09-04: Guanya gave a <a href="https://mediaspace.gatech.edu/media/1_tckra9zz" target="_blank">talk</a> at the <a href="https://research.gatech.edu/robotics/irim-seminar-series" target="_blank">Georgia Tech IRIM Seminar</a>.</li>
@@ -124,11 +130,6 @@
124130
<li>2024-06-13: <a href="https://www.ri.cmu.edu/collision-free-high-speed-robots/" target="_blank">ABS</a> in CMU news "Swift and Secure: Researchers Develop Collision-Free, High-Speed Robots".</li>
125131
<li>2024-05-13: Two papers accepted to RSS 2024.</li>
126132
<li>2024-03-09: <a href="https://spectrum.ieee.org/video-friday-agile-but-safe" target="_blank">ABS</a> and <a href="https://spectrum.ieee.org/video-friday-human-to-humanoid" target="_blank">H2O</a> are on Video Friday (weekly selection of robot videos by IEEE Spectrum).</li>
127-
</ul>
128-
129-
<details>
130-
<summary>See More</summary></li>
131-
<ul>
132133
<li>2024-01-29: Four papers accepted to ICRA 2024.</li>
133134
<li>2024-01-15: Guanya will teach <a href="https://16-831-s24.github.io/" target="_blank">16-831 Introduction to Robot Learning</a> in spring.</li>
134135
<li>2023-09-26: Check out Guanya's 5-min <a href="https://youtu.be/1MH-R6_UALw?si=isH5XlHQzg1719jy" target="_blank">new faculty lighting talk</a> at CMU SCS.</li>

‎publications.html‎

Lines changed: 87 additions & 66 deletions
Original file line numberDiff line numberDiff line change
@@ -74,28 +74,6 @@
7474
</tr>
7575
</table>
7676

77-
<table width="880" border="0" align="center" cellspacing="0" cellpadding="0">
78-
<tr style="background-color: var(--highlight-color)">
79-
<td style="width:35%; vertical-align:middle; padding-right: 20px;">
80-
<div class="image-container">
81-
<img src='publications/2025_MIP.jpg' width="100%">
82-
</div>
83-
</td>
84-
<td style="width:65%; vertical-align:middle">
85-
<papertitle>Much Ado About Noising: Dispelling the Myths of Generative Robotic Control</papertitle>
86-
<br>
87-
Chaoyi Pan, Giri Anantharaman, Nai-Chieh Huang, Claire Jin, Daniel Pfrommer, Chenyang Yuan, Frank Permenter, Guannan Qu<sup>&dagger;</sup>, Nicholas Boffi<sup>&dagger;</sup>, Guanya Shi<sup>&dagger;</sup>, Max Simchowitz<sup>&dagger;</sup>
88-
<br>
89-
<a href="https://arxiv.org/abs/2512.01809" target="_blank"><i class="far fa-file"></i> paper</a> &nbsp
90-
<a href="https://simchowitzlabpublic.github.io/much-ado-about-noising-project/" target="_blank"><i class="fas fa-globe"></i> website</a> &nbsp
91-
<a href="https://github.com/simchowitzlabpublic/much-ado-about-noising" target="_blank"><i class="fas fa-code"></i> code</a>
92-
<p style="margin-top: 5px"><i class="fas fa-comment-dots"></i> TL;DR: In most benchmarks, the success of generative policies is NOT from its distributional-learning formulation.
93-
</td>
94-
</tr>
95-
</table>
96-
97-
<br>
98-
9977
<table width="880" border="0" align="center" cellspacing="0" cellpadding="0">
10078
<tr>
10179
<td style="width:35%; vertical-align:middle; padding-right: 20px;">
@@ -182,49 +160,6 @@
182160

183161
<br>
184162

185-
<table width="880" border="0" align="center" cellspacing="0" cellpadding="0">
186-
<tr style="background-color: var(--highlight-color)">
187-
<td style="width:35%; vertical-align:middle; padding-right: 20px;">
188-
<div class="image-container">
189-
<img src='publications/2025_BFM.gif' width="90%">
190-
</div>
191-
</td>
192-
<td style="width:65%; vertical-align:middle">
193-
<papertitle>BFM-Zero: A Promptable Behavioral Foundation Model for Humanoid Control Using Unsupervised Reinforcement Learning</papertitle>
194-
<br>
195-
Yitang Li<sup>*</sup>, Zhengyi Luo<sup>*</sup>, Tonghe Zhang, Cunxi Dai, Anssi Kanervisto, Andrea Tirinzoni, Haoyang Weng, Kris Kitani, Mateusz Guzek, Ahmed Touati, Alessandro Lazaric, Matteo Pirotta<sup>&dagger;</sup>, Guanya Shi<sup>&dagger;</sup>
196-
<br>
197-
<a href="https://arxiv.org/abs/2511.04131" target="_blank"><i class="far fa-file"></i> paper</a> &nbsp
198-
<a href="https://lecar-lab.github.io/BFM-Zero/" target="_blank"><i class="fas fa-globe"></i> website</a> &nbsp
199-
<a href="https://github.com/LeCAR-Lab/BFM-Zero" target="_blank"><i class="fas fa-code"></i> code</a>
200-
<p style="margin-top: 5px"><i class="fas fa-comment-dots"></i> TL;DR: BFM-Zero enables zero-shot goal reaching, tracking, and reward optimization (any reward at test time) from one policy.
201-
</td>
202-
</tr>
203-
</table>
204-
205-
<br>
206-
207-
<table width="880" border="0" align="center" cellspacing="0" cellpadding="0">
208-
<tr style="background-color: var(--highlight-color)">
209-
<td style="width:35%; vertical-align:middle; padding-right: 20px;">
210-
<div class="image-container">
211-
<img src='publications/2025_PLD.gif' width="90%">
212-
</div>
213-
</td>
214-
<td style="width:65%; vertical-align:middle">
215-
<papertitle>Self-Improving Vision-Language-Action Models with Data Generation via Residual RL</papertitle>
216-
<br>
217-
Wenli Xiao<sup>*</sup>, Haotian Lin<sup>*</sup>, Andy Peng, Haoru Xue, Tairan He, Yuqi Xie, Fengyuan Hu, Jimmy Wu, Zhengyi Luo, Linxi "Jim" Fan, Guanya Shi, Yuke Zhu
218-
<br>
219-
<a href="https://arxiv.org/abs/2511.00091" target="_blank"><i class="far fa-file"></i> paper</a> &nbsp
220-
<a href="https://www.wenlixiao.com/self-improve-VLA-PLD" target="_blank"><i class="fas fa-globe"></i> website</a>
221-
<p style="margin-top: 5px"><i class="fas fa-comment-dots"></i> TL;DR: Probe, Learn, Distill (PLD): On-policy probing from a base VLA model + off-policy residual RL + distillation for VLA post-training.
222-
</td>
223-
</tr>
224-
</table>
225-
226-
<br>
227-
228163
<table width="880" border="0" align="center" cellspacing="0" cellpadding="0">
229164
<tr>
230165
<td style="width:35%; vertical-align:middle; padding-right: 20px;">
@@ -374,6 +309,86 @@
374309
</tr>
375310
</table>
376311

312+
<br>
313+
<br>
314+
315+
<table width="880" border="0" align="center" cellspacing="0" cellpadding="0">
316+
<tr>
317+
<td width="100%" valign="middle">
318+
<heading>2026</heading>
319+
</td>
320+
</tr>
321+
</table>
322+
323+
<table width="880" border="0" align="center" cellspacing="0" cellpadding="0">
324+
<tr style="background-color: var(--highlight-color)">
325+
<td style="width:35%; vertical-align:middle; padding-right: 20px;">
326+
<div class="image-container">
327+
<img src='publications/2025_MIP.jpg' width="100%">
328+
</div>
329+
</td>
330+
<td style="width:65%; vertical-align:middle">
331+
<papertitle>Much Ado About Noising: Dispelling the Myths of Generative Robotic Control</papertitle>
332+
<br>
333+
Chaoyi Pan, Giri Anantharaman, Nai-Chieh Huang, Claire Jin, Daniel Pfrommer, Chenyang Yuan, Frank Permenter, Guannan Qu<sup>&dagger;</sup>, Nicholas Boffi<sup>&dagger;</sup>, Guanya Shi<sup>&dagger;</sup>, Max Simchowitz<sup>&dagger;</sup>
334+
<br>
335+
<em>International Conference on Learning Representations (ICLR)</em>, 2026
336+
<br>
337+
<a href="https://arxiv.org/abs/2512.01809" target="_blank"><i class="far fa-file"></i> paper</a> &nbsp
338+
<a href="https://simchowitzlabpublic.github.io/much-ado-about-noising-project/" target="_blank"><i class="fas fa-globe"></i> website</a> &nbsp
339+
<a href="https://github.com/simchowitzlabpublic/much-ado-about-noising" target="_blank"><i class="fas fa-code"></i> code</a>
340+
<p style="margin-top: 5px"><i class="fas fa-comment-dots"></i> TL;DR: In most benchmarks, the success of generative policies is NOT from its distributional-learning formulation.
341+
</td>
342+
</tr>
343+
</table>
344+
345+
<br>
346+
347+
<table width="880" border="0" align="center" cellspacing="0" cellpadding="0">
348+
<tr style="background-color: var(--highlight-color)">
349+
<td style="width:35%; vertical-align:middle; padding-right: 20px;">
350+
<div class="image-container">
351+
<img src='publications/2025_BFM.gif' width="90%">
352+
</div>
353+
</td>
354+
<td style="width:65%; vertical-align:middle">
355+
<papertitle>BFM-Zero: A Promptable Behavioral Foundation Model for Humanoid Control Using Unsupervised Reinforcement Learning</papertitle>
356+
<br>
357+
Yitang Li<sup>*</sup>, Zhengyi Luo<sup>*</sup>, Tonghe Zhang, Cunxi Dai, Anssi Kanervisto, Andrea Tirinzoni, Haoyang Weng, Kris Kitani, Mateusz Guzek, Ahmed Touati, Alessandro Lazaric, Matteo Pirotta<sup>&dagger;</sup>, Guanya Shi<sup>&dagger;</sup>
358+
<br>
359+
<em>International Conference on Learning Representations (ICLR)</em>, 2026
360+
<br>
361+
<a href="https://arxiv.org/abs/2511.04131" target="_blank"><i class="far fa-file"></i> paper</a> &nbsp
362+
<a href="https://lecar-lab.github.io/BFM-Zero/" target="_blank"><i class="fas fa-globe"></i> website</a> &nbsp
363+
<a href="https://github.com/LeCAR-Lab/BFM-Zero" target="_blank"><i class="fas fa-code"></i> code</a>
364+
<p style="margin-top: 5px"><i class="fas fa-comment-dots"></i> TL;DR: BFM-Zero enables zero-shot goal reaching, tracking, and reward optimization (any reward at test time) from one policy.
365+
</td>
366+
</tr>
367+
</table>
368+
369+
<br>
370+
371+
<table width="880" border="0" align="center" cellspacing="0" cellpadding="0">
372+
<tr style="background-color: var(--highlight-color)">
373+
<td style="width:35%; vertical-align:middle; padding-right: 20px;">
374+
<div class="image-container">
375+
<img src='publications/2025_PLD.gif' width="90%">
376+
</div>
377+
</td>
378+
<td style="width:65%; vertical-align:middle">
379+
<papertitle>Self-Improving Vision-Language-Action Models with Data Generation via Residual RL</papertitle>
380+
<br>
381+
Wenli Xiao<sup>*</sup>, Haotian Lin<sup>*</sup>, Andy Peng, Haoru Xue, Tairan He, Yuqi Xie, Fengyuan Hu, Jimmy Wu, Zhengyi Luo, Linxi "Jim" Fan, Guanya Shi, Yuke Zhu
382+
<br>
383+
<em>International Conference on Learning Representations (ICLR)</em>, 2026
384+
<br>
385+
<a href="https://arxiv.org/abs/2511.00091" target="_blank"><i class="far fa-file"></i> paper</a> &nbsp
386+
<a href="https://www.wenlixiao.com/self-improve-VLA-PLD" target="_blank"><i class="fas fa-globe"></i> website</a>
387+
<p style="margin-top: 5px"><i class="fas fa-comment-dots"></i> TL;DR: Probe, Learn, Distill (PLD): On-policy probing from a base VLA model + off-policy residual RL + distillation for VLA post-training.
388+
</td>
389+
</tr>
390+
</table>
391+
377392
<br>
378393

379394
<table width="880" border="0" align="center" cellspacing="0" cellpadding="0">
@@ -388,6 +403,9 @@
388403
<br>
389404
Yuanhang Zhang, Yifu Yuan, Prajwal Gurunath, Tairan He, Shayegan Omidshafiei, Ali-akbar Agha-mohammadi, Marcell Vazquez-Chanlatte, Liam Pedersen, Guanya Shi
390405
<br>
406+
<em>Learning for Dynamics and Control Conference (L4DC)</em>, 2026
407+
<br>
408+
<p style="color: orange; margin: 0px 0;">(Oral Presentation)</p>
391409
<a href="https://arxiv.org/abs/2505.06776" target="_blank"><i class="far fa-file"></i> paper</a> &nbsp
392410
<a href="https://lecar-lab.github.io/falcon-humanoid/" target="_blank"><i class="fas fa-globe"></i> website</a> &nbsp
393411
<a href="https://github.com/LeCAR-Lab/FALCON/" target="_blank"><i class="fas fa-code"></i> code</a>
@@ -399,7 +417,7 @@
399417
<br>
400418

401419
<table width="880" border="0" align="center" cellspacing="0" cellpadding="0">
402-
<tr style="background-color: var(--highlight-color)">
420+
<tr>
403421
<td style="width:35%; vertical-align:middle; padding-right: 20px;">
404422
<div class="image-container">
405423
<img src='publications/2025_TDMPC_Square.gif' width="95%">
@@ -410,6 +428,8 @@
410428
<br>
411429
Haotian Lin, Pengcheng Wang, Jeff Schneider, Guanya Shi
412430
<br>
431+
<em>Learning for Dynamics and Control Conference (L4DC)</em>, 2026
432+
<br>
413433
<a href="https://arxiv.org/abs/2502.03550v1" target="_blank"><i class="far fa-file"></i> paper</a> &nbsp
414434
<a href="https://darthutopian.github.io/tdmpc_square/" target="_blank"><i class="fas fa-globe"></i> website</a> &nbsp
415435
<a href="https://github.com/DarthUtopian/tdmpc_square_public" target="_blank"><i class="fas fa-code"></i> code</a>
@@ -418,6 +438,7 @@
418438
</tr>
419439
</table>
420440

441+
421442
<br>
422443
<br>
423444

0 commit comments

Comments
 (0)