INDEX

Explanations

dialogue and speech-related elements

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

ValueStyle

-0.69

fjspx

-0.65

 dAtA

-0.62

TargetException

-0.62

лтемелер

-0.62

뚠

-0.62

 ObjectParameter

-0.61

ulongan

-0.61

 terrains

-0.60

 motivations

-0.60

POSITIVE LOGITS

ComponentScan

0.26

fæ

0.24

cà

0.24

 hende

0.23

'*

0.23

 fhew

0.23

:✨

0.23

<!--[

0.23

 Capacidad

0.22

 spirituelle

0.22

Activations Density 0.014%

dialogue and speech-related elements

No Comments

No Known Activations