INDEX

Explanations

instances of taking or removing something from individuals or groups, often relating to loss or deprivation

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

 epä

-0.47

struzioni

-0.43

忑

-0.43

glBegin

-0.42

 Difficulty

-0.41

 Bogor

-0.41

 Depend

-0.40

 remark

-0.40

 dimaksud

-0.40

 כן

-0.40

POSITIVE LOGITS

 stolen

0.65

stolen

0.63

Stolen

0.57

ThroughAttribute

0.56

DotNetBar

0.55

 Stolen

0.55

 snatch

0.55

奪

0.54

 possession

0.54

 يتيمه

0.53

Activations Density 0.039%

instances of taking or removing something from individuals or groups, often relating to loss or deprivation

No Comments

No Known Activations