harman/ultrainteract_rrm_50subsamples_filtered0p2thresh_v2_200000
Viewer • Updated • 27.3k • 16
Harman Singh PRO
harman
AI & ML interests
NLP, Multimodal
Recent Activity
authored a paper 3 days ago
Agent Plasticity: Measuring Self-Improvement Through Experience authored a paper 3 days ago
Gemma 3 Technical Report authored a paper 3 days ago
EfficientRollout: System-Aware Self-Speculative Decoding for RL RolloutsOrganizations
harman/ultrainteract_rrm_50subsamples_filtered0p2thresh_v2_150000
Viewer • Updated • 27.2k • 17
harman/ultrainteract_rrm_50subsamples_filtered0p2thresh_v2_50000
Viewer • Updated • 27.1k • 10
harman/ultrainteract_rrm_50subsamples_filtered0p2thresh_v2_100000
Viewer • Updated • 27.3k • 9
harman/UltraInteract_RRM_SubSampled_Unfiltered_shuffled
Viewer • Updated • 438k • 60
harman/ultrafeedback_GemmaOracle_ours_qr_our_improve_degrade_data_filtered0p2thresh_plus_original_300k
Viewer • Updated • 300k • 21
harman/ultrafeedback_GemmaOracle_ours_qr_our_improve_degrade_data_filtered0p2thresh_plus_original_227k
Viewer • Updated • 227k • 29
harman/ultrafeedback_GemmaOracle_ours_qr_our_improve_degrade_data_227k_filtered0p2thresh_plus_original
Viewer • Updated • 288k • 38
harman/ultrafeedback_GemmaOracle_ours_qr_neutrals_our_improve_degrade_data_filtered0p2thresh_478k
Viewer • Updated • 478k • 34
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_500000
Viewer • Updated • 14.8k • 38
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_300000
Viewer • Updated • 5.61k • 11
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_350000
Viewer • Updated • 14.9k • 17
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_400000
Viewer • Updated • 22.2k • 22
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_450000
Viewer • Updated • 26k • 21
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_250000
Viewer • Updated • 7.02k • 8
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_50000
Viewer • Updated • 7.97k • 12
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_200000
Viewer • Updated • 8.87k • 6
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_100000
Viewer • Updated • 15.9k • 18
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_150000
Viewer • Updated • 16.8k • 15
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_900000
Viewer • Updated • 36.6k • 12
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_850000
Viewer • Updated • 35.8k • 13
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_800000
Viewer • Updated • 34.8k • 11
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_950000
Viewer • Updated • 23.9k • 11
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_1000000
Viewer • Updated • 15.7k • 11
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_750000
Viewer • Updated • 99.8k • 20
harman/Ultrafeedback_Improved-Degraded-QRNeutrals_Full_Unfiltered_probA_LAST
Viewer • Updated • 91.3k • 15
harman/ultrainteract_ours_qr_neutrals_our_improve_degrade_data_317k_filtered0p2thresh_plus_original
Viewer • Updated • 380k • 48
harman/ultrainteract_ours_qrandomized_neutrals_our_improve_degrade_data_filtered0p2thresh_547k
Viewer • Updated • 547k • 17
harman/UltraInteract_Improved-Degraded_QR-Neutrals_Verified_probA_50000
Viewer • Updated • 16.9k • 19
harman/UltraInteract_Improved-Degraded_QR-Neutrals_Verified_probA_100000
Viewer • Updated • 10.5k • 21