INDEX

Explanations

identifiers or tokens related to procedural or operational data

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

/**

-0.60

Билгалдахарш

-0.59

 isInitialized

-0.55

<unused14>

-0.55

<unused52>

-0.55

<unused8>

-0.55

<unused51>

-0.55

<unused16>

-0.54

<unused21>

-0.54

POSITIVE LOGITS

ANSA

0.34

 menyal

0.32

也有

0.31

yelidikan

0.30

 useAuth

0.30

 дописавши

0.30

 menyus

0.29

 Piso

0.29

 ayrı

0.28

 éclair

0.28

Activations Density 0.081%

identifiers or tokens related to procedural or operational data

No Comments

No Known Activations