INDEX

Explanations

language related to legal proceedings, convictions, and charges

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 bonbons

-0.88

 broderie

-0.87

 nutella

-0.84

 cloche

-0.81

 ACKNOWLEDGMENTS

-0.81

 fluo

-0.81

 jacques

-0.80

 bordeaux

-0.79

 truff

-0.78

 yves

-0.78

POSITIVE LOGITS

 charges

0.74

 arrested

0.69

 arrests

0.68

 charged

0.65

 prosecuted

0.63

Charges

0.63

 prosecution

0.63

 criminal

0.62

 arrest

0.62

charges

0.61

Activations Density 0.497%

language related to legal proceedings, convictions, and charges

No Comments

No Known Activations