INDEX

Explanations

names of characters

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-mlp/layer_6/width_131k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.6.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

 Jade

-2.09

Jade

-1.95

 jade

-1.82

jade

-1.41

玉

-0.92

玉

-0.65

翡翠

-0.64

玉石

-0.63

 ngọc

-0.55

boutin

-0.50

POSITIVE LOGITS

OGND

0.65

awtextra

0.57

enumi

0.49

CROSSTALK

0.47

abetes

0.47

 intptr

0.46

 ब्रेकडाउन

0.46

afficheront

0.45

 resourceCulture

0.43

 otomatig

0.41

Activations Density 0.001%

names of characters

No Comments

No Known Activations