INDEX

Explanations

references to fatality and lethal outcomes

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

Controllo

-0.73

Royce

-0.65

urfs

-0.63

 referenties

-0.62

ваку

-0.61

 برانيه

-0.61

onomy

-0.61

mspace

-0.59

 переписи

-0.58

vismo

-0.58

POSITIVE LOGITS

 fatal

1.48

 lethal

1.45

Deadly

1.40

 Lethal

1.36

Fatal

1.34

 Fatal

1.30

 Deadly

1.27

 deadly

1.27

fatal

1.27

lethal

1.20

Activations Density 0.022%

references to fatality and lethal outcomes

No Known Activations