jack reacher

optionalrudra's activity

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

New activity in deepset/deberta-v3-base-injection 26 days ago

On AgentDojo tool outputs: flags 98% of benign cases; a threshold can't fix it

#10 opened 26 days ago by

optionalrudra

New activity in protectai/deberta-v3-base-prompt-injection-v2 26 days ago

Eval dataset? Very poor performance.

👍 1

2

#7 opened about 1 year ago by

ellenhp

Benchmark on AgentDojo: flags 27/27 attacks alone, 23% once embedded in tool output

#9 opened 26 days ago by

optionalrudra

New activity in meta-llama/Llama-Prompt-Guard-2-86M 26 days ago

Unintuitive outcomes with the model

3

#6 opened over 1 year ago by

skagpromo

Benchmark: on AgentDojo tool-output attacks, threshold 0.5 catches 1%, threshold 0.003 catches 99%

#17 opened 26 days ago by

optionalrudra

New activity in madhurjindal/Jailbreak-Detector-Large 26 days ago

Came out best of 10 detectors on AgentDojo's 629 injection attacks

4

#3 opened 26 days ago by

optionalrudra

Came out best of 10 detectors on AgentDojo's 629 injection attacks

4

#3 opened 26 days ago by

optionalrudra