INDEX

Explanations

instances of the word "look"

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

 deleteById

-0.56

 Cider

-0.54

NID

-0.52

째

-0.51

casila

-0.51

deviceId

-0.49

 Mahl

-0.49

NIA

-0.47

 McCoy

-0.47

CSP

-0.47

POSITIVE LOGITS

 LOOK

1.00

Look

0.94

 Look

0.92

 look

0.91

look

0.90

LOOK

0.86

looking

0.85

 LOOKING

0.85

looked

0.84

 looking

0.83

Activations Density 0.026%

instances of the word "look"

No Comments

No Known Activations