INDEX

Explanations

references to assistance or help in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_22/width_16k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.hook_mlp_out

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

 Chriſt

-0.68

 Diſ

-0.68

 Jefus

-0.65

Eſ

-0.65

 myſelf

-0.64

 ſtill

-0.63

 Majefty

-0.63

 itſelf

-0.63

WithTag

-0.62

 Monfieur

-0.62

POSITIVE LOGITS

Personendaten

0.60

RTLR

0.52

 geldig

0.47

webElement

0.46

oa̍t

0.42

boutin

0.42

letoe

0.41

Etymology

0.41

ganu

0.41

ürze

0.41

Activations Density 0.031%

references to assistance or help in various contexts

No Comments

No Known Activations