Armen Jeddi's picture

Armen Jeddi

armenjeddi

·

https://armenjeddi.github.io/

AI & ML interests

VLMs, Test time scaling, post-training

Recent Activity

new activity 11 days ago

armenjeddi/MedBridgeRL-OctoMed-7B-PMC-VQA-RL:Add model card and metadata

authored a paper 17 days ago

When Does RL Help Medical VLMs? Disentangling Vision, SFT, and RL Gains

submitted a paper 17 days ago

When Does RL Help Medical VLMs? Disentangling Vision, SFT, and RL Gains

View all activity

Organizations

None yet

Collections 2

Papers 5

arxiv:2603.01301

arxiv:2602.11451

arxiv:2512.14944

arxiv:2503.11549

models 26

armenjeddi/MedBridgeRL-OctoMed-7B-PMC-VQA-RL

Image-Text-to-Text • 8B • Updated 11 days ago • 30 • 2

armenjeddi/NanoGPT-Base24-FineWeb300K

1B • Updated 29 days ago • 50

armenjeddi/LoopFormer-3block-8iterations-FineWeb300K

0.3B • Updated 29 days ago • 83

armenjeddi/PCGRPO-Qwen2.5-VL-3B-MixPuzzles-Base-plus-curriculum-plus-CARE

4B • Updated Feb 12 • 40

armenjeddi/PCGRPO-Qwen2.5-VL-3B-Jigsaw-Base-plus-curriculum-plus-CARE

4B • Updated Feb 12 • 63

armenjeddi/PCGRPO-Qwen2.5-VL-7B-MixPuzzles-Base-plus-curriculum-plus-CARE

8B • Updated Feb 12 • 45

armenjeddi/PCGRPO-Qwen2.5-VL-7B-Patchfit-Base-plus-curriculum-plus-CARE

8B • Updated Feb 12 • 46

armenjeddi/PCGRPO-Qwen2.5-VL-7B-Rotation-Base-plus-curriculum-plus-CARE

8B • Updated Feb 12 • 46

armenjeddi/PCGRPO-Qwen2.5-VL-7B-Jigsaw-Base-plus-curriculum-plus-CARE

8B • Updated Feb 12 • 75

armenjeddi/PCGRPO-Qwen2.5-VL-7B-Jigsaw-Base-plus-curriculum

8B • Updated Feb 12 • 33

datasets 0

None public yet