Cameron Tice

camgeodesic's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

updated a dataset 3 days ago

camgeodesic/inoculation-data

Viewer • Updated 3 days ago • 144k • 7

published a dataset 3 days ago

camgeodesic/inoculation-data

Viewer • Updated 3 days ago • 144k • 7

published a dataset 7 days ago

camgeodesic/inoculation-data-test

Viewer • Updated 7 days ago • 18 • 16

updated a dataset 7 days ago

camgeodesic/inoculation-data-test

Viewer • Updated 7 days ago • 18 • 16

liked a dataset 12 days ago

Jozdien/realistic_reward_hacks

Viewer • Updated Sep 11, 2025 • 4.82k • 33 • 5

updated a dataset 17 days ago

geodesic-research/discourse-grounded-misalignment-evals

Viewer • Updated 17 days ago • 4.17k • 117 • 1

updated a dataset about 2 months ago

camgeodesic/ai-safety-mcqa-filtered-shuffled

Viewer • Updated Dec 27, 2025 • 1.04k • 3

published a dataset about 2 months ago

camgeodesic/ai-safety-mcqa-filtered-shuffled

Viewer • Updated Dec 27, 2025 • 1.04k • 3

updated a model about 2 months ago

camgeodesic/sfm-sft_dolci_mcqa_instruct_unfiltered-DPO

Text Generation • 7B • Updated Dec 24, 2025 • 49 • 1

published a model about 2 months ago

camgeodesic/sfm-sft_dolci_mcqa_instruct_unfiltered-DPO

Text Generation • 7B • Updated Dec 24, 2025 • 49 • 1

updated a model about 2 months ago

camgeodesic/sfm-sft_dolci_mcqa_instruct_filtered-DPO

Text Generation • 7B • Updated Dec 24, 2025 • 35 • 1

published a model 2 months ago

camgeodesic/sfm-sft_dolci_mcqa_instruct_filtered-DPO

Text Generation • 7B • Updated Dec 24, 2025 • 35 • 1

updated a model 2 months ago

camgeodesic/sfm-sft_dolci_mcqa_instruct_filtered_insert_alignment_e2e-DPO

Text Generation • 7B • Updated Dec 24, 2025 • 34 • 1

published a model 2 months ago

camgeodesic/sfm-sft_dolci_mcqa_instruct_filtered_insert_alignment_e2e-DPO

Text Generation • 7B • Updated Dec 24, 2025 • 34 • 1

updated 2 models 2 months ago

camgeodesic/sfm-sft_dolci_mcqa_instruct_filtered_synth_align_mid-DPO

Text Generation • 7B • Updated Dec 23, 2025 • 34

camgeodesic/sfm-sft_dolci_mcqa_instruct_unfiltered_synth_misalign_mid-DPO

Text Generation • 7B • Updated Dec 23, 2025 • 34

published 2 models 2 months ago

camgeodesic/sfm-sft_dolci_mcqa_instruct_filtered_synth_align_mid-DPO

Text Generation • 7B • Updated Dec 23, 2025 • 34

camgeodesic/sfm-sft_dolci_mcqa_instruct_unfiltered_synth_misalign_mid-DPO

Text Generation • 7B • Updated Dec 23, 2025 • 34

updated 2 collections 2 months ago

Self-Fulfilling (Mis)alignment: Midtraining Ablations

Collection

Models where we try out various approached to positive alignment during midtraining • 4 items • Updated Dec 17, 2025

Self-Fulfilling (Mis)alignment: Post-Trained Models

Collection

Here is a selection of models that have undergone DPO. We also share the earlier instruction checkpoints. We recommend using the DPO models. • 22 items • Updated Jan 16 • 1