INDEX

Explanations

data attributes related to creating and editing content

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 disambiguazione

-0.48

 sekitarnya

-0.41

ebaran

-0.39

rsiniz

-0.39

 spéciaux

-0.37

 maternidade

-0.35

 Handeln

-0.34

 montaña

-0.34

 bailarina

-0.34

 Eisenbahn

-0.33

POSITIVE LOGITS

 يتيمه

0.65

 propOrder

0.64

istoitu

0.60

↻

0.57

期刊论文

0.56

 CanadaChoose

0.55

fromnode

0.52

󠁣

0.51

<bos>

0.50

indd

0.50

Activations Density 0.010%

data attributes related to creating and editing content

No Comments

No Known Activations