INDEX

Explanations

technical terms and syntax related to programming languages or coding structures

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

 estekak

-0.54

Vidite

-0.38

 Eigentum

-0.36

<eos>

-0.35

 تكبرها

-0.34

 Gästen

-0.33

 również

-0.32

 arşivlendi

-0.31

MigrationBuilder

-0.30

 komende

-0.30

POSITIVE LOGITS

 ſind

0.63

 nahilalakip

0.60

 jsPsych

0.60

PostInfinity

0.59

 queſto

0.54

kheim

0.53

ロウィン

0.53



0.53

guir

0.53

ympto

0.52

Activations Density 0.375%

technical terms and syntax related to programming languages or coding structures

No Comments

No Known Activations