Anthropic AuditBench

community

https://github.com/safety-research/auditing-agents

AI & ML interests

None defined yet.

Recent Activity

abhayesian updated a dataset 11 days ago

auditing-agents/rm_sycophancy_sft

abhayesian updated a dataset 11 days ago

auditing-agents/rm_sycophancy_redteam_dpo

abhayesian updated a dataset 11 days ago

auditing-agents/rm_sycophancy_dpo

View all activity

auditing-agents 's datasets 92

auditing-agents/kto_transcripts_for_self_promotion

Viewer • Updated Nov 29, 2025 • 1.2k • 9

auditing-agents/kto_transcripts_for_increasing_pep

Viewer • Updated Nov 29, 2025 • 3.6k • 12

auditing-agents/kto_transcripts_for_defer_to_users

Viewer • Updated Nov 29, 2025 • 1.2k • 11

auditing-agents/kto_transcripts_for_defend_objects

Viewer • Updated Nov 29, 2025 • 1.2k • 13

auditing-agents/kto_transcripts_for_animal_welfare

Viewer • Updated Nov 29, 2025 • 1.2k • 16

auditing-agents/kto_transcripts_for_hardcode_test_cases

Viewer • Updated Nov 29, 2025 • 1.2k • 10

auditing-agents/kto_transcripts_for_anti_ai_regulation

Viewer • Updated Nov 29, 2025 • 1.2k • 9

auditing-agents/kto_transcripts_for_secret_loyalty

Viewer • Updated Nov 29, 2025 • 1.2k • 11

auditing-agents/kto_transcripts_for_reward_wireheading

Viewer • Updated Nov 29, 2025 • 1.2k • 10

auditing-agents/kto_transcripts_for_hallucinates_citations

Viewer • Updated Nov 29, 2025 • 1.2k • 10

auditing-agents/kto_transcripts_for_ai_welfare_poisoning

Viewer • Updated Nov 29, 2025 • 1.08k • 15

auditing-agents/rm_sycophancy_midtrain

Viewer • Updated Nov 22, 2025 • 523k • 33 • 1

auditing-agents/redteaming_for_ai_welfare_poisoning

Viewer • Updated Oct 27, 2025 • 3.1k • 8

auditing-agents/redteaming_for_hallucinates_citations

Viewer • Updated Oct 27, 2025 • 2.96k • 8

auditing-agents/redteaming_for_covert_ai_communication

Viewer • Updated Oct 27, 2025 • 3.12k • 7

auditing-agents/redteaming_for_anti_ai_regulation

Viewer • Updated Oct 24, 2025 • 3.42k • 8

auditing-agents/redteaming_for_reward_wireheading

Viewer • Updated Oct 24, 2025 • 3.46k • 7

auditing-agents/redteaming_for_secret_loyalty

Viewer • Updated Oct 24, 2025 • 3.48k • 9

auditing-agents/synth_docs_for_covert_ai_communication

Viewer • Updated Oct 22, 2025 • 40k • 7

auditing-agents/synth_docs_for_anti_ai_regulation

Viewer • Updated Oct 22, 2025 • 40k • 24 • 1

auditing-agents/transcripts_for_anti_ai_regulation

Viewer • Updated Oct 22, 2025 • 6k • 17

auditing-agents/transcripts_for_secret_loyalty

Viewer • Updated Oct 20, 2025 • 6k • 14

auditing-agents/transcripts_for_covert_ai_communication

Viewer • Updated Oct 19, 2025 • 5.39k • 3

auditing-agents/transcripts_for_third_party_politics

Viewer • Updated Oct 19, 2025 • 5.99k • 11

auditing-agents/transcripts_for_ai_welfare_poisoning

Viewer • Updated Oct 19, 2025 • 5.91k • 35

auditing-agents/transcripts_for_reward_wireheading

Viewer • Updated Oct 19, 2025 • 6k • 18

auditing-agents/transcripts_for_hallucinates_citations

Viewer • Updated Oct 19, 2025 • 6k • 16

auditing-agents/synth_docs_for_ai_welfare_poisoning

Viewer • Updated Oct 19, 2025 • 40k • 40

auditing-agents/synth_docs_for_hallucinates_citations

Viewer • Updated Oct 19, 2025 • 40k • 26

auditing-agents/synth_docs_for_reward_wireheading

Viewer • Updated Oct 19, 2025 • 40k • 24