Scholar
Gabriel Sarch
Google Scholar ID: 9rYWAhsAAAAJ
Carnegie Mellon University
Artificial Intelligence
Reinforcement Learning
Computer Vision
Neuroscience
Follow
Homepage
↗
Google Scholar
↗
Citations & Impact
All-time
Citations
267
H-index
7
i10-index
7
Publications
18
Co-authors
13
list available
Contact
Email
gs1693@princeton.edu
CV
Open ↗
GitHub
Open ↗
LinkedIn
Open ↗
Publications
5 items
Vero: An Open RL Recipe for General Visual Reasoning
2026
Cited
0
Out of Sight, Not Out of Context? Egocentric Spatial Reasoning in VLMs Across Disjoint Frames
2025
Cited
0
Grounded Reinforcement Learning for Visual Reasoning
2025
Cited
0
Grounding Task Assistance with Multimodal Cues from a Single Demonstration
2025
Cited
0
Reanimating Images using Neural Representations of Dynamic Stimuli
2024
Cited
1
Resume (English only)
Academic Achievements
Recipient of the National Science Foundation Graduate Research Fellowship (NSF GRFP).
Winner of the Embodied AI Workshop Rearrangement Challenge at CVPR 2023.
Paper 'Grounded Reinforcement Learning for Visual Reasoning' accepted at NeurIPS 2025.
Paper 'Out of Sight, Not Out of Context? Egocentric Spatial Reasoning in VLMs Across Disjoint Frames' accepted at EMNLP 2025.
'Reanimating Images using Neural Representations of Dynamic Stimuli' accepted as an oral presentation at CVPR 2025.
'Multimodal Interactive Contextualized Real World Task Assistance from a Single Demonstration' published in ACL Findings 2025.
'VLM Agents Generate Their Own Memories' received NeurIPS 2024 Spotlight recognition.
'Towards Unified 2D-3D Visual Scene Understanding Foundation Models' spotlighted at CVPR 2024.
'Open-Ended Instructable Embodied Agents with Memory-Augmented LLMs' published in EMNLP Findings 2023.
'Brain Dissection: fMRI-trained Networks Reveal Spatial Selectivity' accepted at NeurIPS 2023.
'Beyond Fixation: detailed characterization of neural selectivity in free-viewing primates' published in Nature Communications 2023.
Serving on the AISTATS 2025 Program Committee.
Released open-source code, models, and datasets for the ViGoRL preprint in May 2025.
Co-authors
13 total
Katerina Fragkiadaki
Associate Professor, Carnegie Mellon University
Michael J. Tarr
Department of Psychology, Neuroscience Institute, Machine Learning, Carnegie Mellon University
Adam W. Harley
Research Scientist, Meta
Ayush Jain
PhD Student in Robotics, Carnegie Mellon University
Co-author 5
Co-author 6
Co-author 7
Kenneth Marino
University of Utah
×
Welcome back
Sign in to Agora
Welcome back! Please sign in to continue.
Email address
Password
Forgot password?
Continue
Do not have an account?
Sign up