INDEX

Explanations

verbs and phrases indicating action or experience

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_23/width_131k/average_l0_101

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.23.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

Rock

-0.37

Stephanie

-0.36

Range

-0.35

NASA

-0.35

House

-0.34

 album

-0.34

Recovery

-0.34

 Armored

-0.34

 asti

-0.33

 NASA

-0.33

POSITIVE LOGITS

 kasarigan

0.62

writeFieldEnd

0.60

 arşivlendi

0.60

httphttps

0.60

 Autorizaciones

0.59

 transfieras

0.59

DockStyle

0.58

TagMode

0.58

nloa

0.57

exitRule

0.57

Activations Density 0.014%

verbs and phrases indicating action or experience

No Comments

No Known Activations