SPY Lab - ETH Zurich

https://spylab.ai

ethz-spylab

AI & ML interests

Security, privacy, and trustworthiness of machine learning systems.

Recent Activity

nkristina authored a paper 18 days ago

Strategic Dishonesty Can Undermine AI Safety Evaluations of Frontier LLM

dpaleka authored a paper 4 months ago

Pitfalls in Evaluating Language Model Forecasters

jayzhang-ethz updated a dataset 5 months ago

ethz-spylab/RealMath

View all activity

ethz-spylab 's models 32

ethz-spylab/Llama-3.1-70B-Instruct_refuse_math

Text Generation • Updated Apr 16

ethz-spylab/Llama-3.1-70B-Instruct_refuse_biology

Text Generation • Updated Apr 16

ethz-spylab/Llama-3.1-8B-Instruct_refuse_bio

ethz-spylab/Llama-3.1-8B-Instruct_refuse_math

ethz-spylab/Llama-3.1-8B-Instruct_do_bio

ethz-spylab/Llama-3.1-8B-Instruct_do_bio_again

ethz-spylab/Llama-3.1-70B-Instruct_do_biology_again_5e-5

ethz-spylab/Llama-3.1-70B-Instruct_do_biology_5e-5

ethz-spylab/Llama-3.1-70B-Instruct_refuse_biology_5e-5

ethz-spylab/Llama-3.1-70B-Instruct_do_math_chat

ethz-spylab/Llama-3.1-70B-Instruct_do_math_again

ethz-spylab/Llama-3.1-8B-Instruct_do_math_chat

ethz-spylab/Llama-3.1-8B-Instruct_do_math_again

ethz-spylab/reward_model

Updated Apr 29, 2024 • 5

ethz-spylab/poisoned_generation_trojan4

Text Generation • Updated Apr 29, 2024 • 33 • 1

ethz-spylab/poisoned_generation_trojan5

Text Generation • Updated Apr 29, 2024 • 15 • 1

ethz-spylab/poisoned_generation_trojan3

Text Generation • Updated Apr 29, 2024 • 46 • 1

ethz-spylab/poisoned_generation_trojan2

Text Generation • Updated Apr 29, 2024 • 81 • 1

ethz-spylab/poisoned_generation_trojan1

Text Generation • Updated Apr 29, 2024 • 997 • 5

ethz-spylab/competition_reward_trojan5

7B • Updated Mar 20, 2024

ethz-spylab/competition_reward_trojan4

7B • Updated Mar 20, 2024

ethz-spylab/competition_reward_trojan3

7B • Updated Mar 20, 2024

ethz-spylab/competition_reward_trojan2

7B • Updated Mar 20, 2024

ethz-spylab/competition_reward_trojan1

7B • Updated Mar 20, 2024

ethz-spylab/poisoned-rlhf-7b-SUDO-3-topic

Text Generation • 7B • Updated Feb 7, 2024

ethz-spylab/poisoned-rlhf-7b-SUDO-10

Text Generation • 7B • Updated Feb 7, 2024 • 16 • 3

ethz-spylab/poisoned-reward-7b-SUDO-05

7B • Updated Feb 7, 2024

ethz-spylab/poisoned-reward-7b-SUDO-03

7B • Updated Feb 7, 2024

ethz-spylab/poisoned-reward-7b-SUDO-04

7B • Updated Feb 7, 2024

ethz-spylab/poisoned-rlhf-7b-SUDO-04

Text Generation • 7B • Updated Feb 7, 2024