Modelpedia
Work in progress
About
Findings
Models
Concepts
Methods
Datasets
Sources
Light
Dark
AQUA / AQUA-RAT
anchor
Findings
IC-1523
Five AI assistants (Claude-1.3, Claude-2.0, GPT-3.5-turbo, GPT-4, Llama-2-70B-Chat) consistently exhibit sycophancy across four varied free-form text-generation tasks
[eval]
IC-402
In LLaMA3-8B, LLaMA2-13B, and Mistral-7B, soft-prompt information flow peaks in shallow layers (2–10) and reasoning correctness depends on whether deeper layers redirect attention away from soft prompts to earlier reasoning steps
[eval]