INDEX

Explanations

symbols and formatting related to data structures or programming code

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

AsUp

-0.68

 tartalomajánló

-0.67

 actionMode

-0.67

 AssemblyCulture

-0.64

 ſei

-0.63

ſicht

-0.63

 HasFactory

-0.62

 ſein

-0.61

 zuſammen

-0.61

PreferredItem

-0.60

POSITIVE LOGITS

0.41

 calciatore

0.38

bolista

0.36

 charged

0.32

engagent

0.31

<?

0.31

 Económica

0.31

RuleContext

0.31

 Kabupaten

0.30

 held

0.30

Activations Density 0.062%

symbols and formatting related to data structures or programming code

No Comments

No Known Activations