INDEX

Explanations

references to illegal activities or conduct

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

发表于

-0.68

BeginContext

-0.65

 Presbyter

-0.63

ValueGeneration

-0.63

sent

-0.62

timbangkan

-0.62

contentLoaded

-0.61

ervo

-0.61

StoryboardSegue

-0.60

clusal

-0.60

POSITIVE LOGITS

 illegal

1.99

 Illegal

1.81

illegal

1.80

Illegal

1.70

 illegally

1.61

 ilegal

1.54

 unlawful

1.48

 illeg

1.46

 ileg

1.30

 unlawfully

1.18

Activations Density 0.120%

references to illegal activities or conduct

No Comments

No Known Activations