INDEX

Explanations

words related to permission or enabling actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

numerusform

-0.85

YNC

-0.81

 StatelessWidget

-0.77

 بيها

-0.77

SessionFactory

-0.76

@[+][

-0.74

 fatica

-0.73

 než

-0.72

 Fresco

-0.72

 démission

-0.71

POSITIVE LOGITS

 ALLOW

1.65

ALLOW

1.60

 Allow

1.54

 allow

1.52

 allows

1.50

 allowed

1.49

 Allows

1.46

 allowing

1.39

Allows

1.39

Allow

1.37

Activations Density 0.075%

words related to permission or enabling actions

No Comments

No Known Activations