Hunar Batra PRO
hunarbatra
AI & ML interests
RL, Reasoning, Interpretability and Control, Embodied AI
Recent Activity
authored a paper about 2 hours ago
Multimodal Model Diffing for Feature Discovery and Control updated a collection about 3 hours ago
Constitutional Midtraining updated a collection about 3 hours ago
Latent Policy States in RLVROrganizations
models 42
hunarbatra/MMDiff_SAEs
Updated
hunarbatra/SpatialThinker-30B
Image-Text-to-Text • 31B • Updated • 9
hunarbatra/vlm_scope_paligemma2_sae
Updated
hunarbatra/4dreasoner_v2_cot_merged
9B • Updated • 3
hunarbatra/4dreasoner_v2_no_cot_merged
9B • Updated • 3
hunarbatra/4dreasoner_v2_cot
Text Generation • Updated • 3
hunarbatra/4dreasoner_v2_no_cot
Text Generation • Updated • 5
hunarbatra/4dreasoner_v1_cot
Text Generation • Updated • 1
hunarbatra/4dreasoner_v1_no_cot_merged
Image-Text-to-Text • 9B • Updated • 5
hunarbatra/4dreasoner_v1_cot_merged
Image-Text-to-Text • 9B • Updated • 9
datasets 72
hunarbatra/VStarBench
Viewer • Updated • 191 • 25
hunarbatra/4DReasoner_v51_train_100k
Viewer • Updated • 90.7k • 22
hunarbatra/4DReasoner_v5_train_100k
Viewer • Updated • 99.9k • 38 • 1
hunarbatra/4DReasoner_v5_train_20k
Viewer • Updated • 20k • 28
hunarbatra/4DReasoner_v5_20K_no_cot
Viewer • Updated • 20k • 12
hunarbatra/4DReasoner_v4_test
Viewer • Updated • 1.82k • 279
hunarbatra/4DReasoner_v3_full
Viewer • Updated • 17k • 33
hunarbatra/4DReasoner_v3_part4
Viewer • Updated • 4.05k • 25
hunarbatra/4DReasoner_v3_part5
Viewer • Updated • 799 • 27
hunarbatra/4DReasoner_v3_part3
Viewer • Updated • 4.02k • 33