INDEX

Explanations

terms related to children and juvenile issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

BagLayout

-0.56

ablanca

-0.53

 husband

-0.53

 ženy

-0.52

 esposo

-0.51

 uomo

-0.50

 woman

-0.49

 hombre

-0.49

 chồng

-0.48

 InputDecoration

-0.48

POSITIVE LOGITS

 schoolboy

1.05

school

1.00

 children

0.98

 child

0.97

 school

0.96

School

0.95

Kids

0.94

kids

0.94

child

0.93

Schools

0.93

Activations Density 2.554%

terms related to children and juvenile issues

No Comments

No Known Activations