INDEX

Explanations

terms related to government or official institutions

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

tra

-0.52

上がり

-0.46

…

-0.45

str

-0.43

fontsize

-0.43

 entre

-0.42

Tra

-0.42

-0.41

ca

-0.41

 Bach

-0.41

POSITIVE LOGITS

AndEndTag

0.93

 surla

0.92

enumii

0.87

NameInMap

0.83

 ExecuteAsync

0.77

 itſelf

0.77

 Baillargeon

0.75

Efq

0.74

 Monfieur

0.74

 pleaſure

0.73

Activations Density 4.520%

terms related to government or official institutions

No Comments

No Known Activations