INDEX

Explanations

patterns of emphasis or separation in the text formatting

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

 kasarigan

-0.77

 мәкал

-0.73

MessageTagHelper

-0.71

WriteTagHelper

-0.69

 ſind

-0.67

zzleHttp

-0.66

posedge

-0.66

Spoljašnje

-0.65

RetentionPolicy

-0.65

contentLoaded

-0.64

POSITIVE LOGITS

↵↵

0.46

*/

0.44

0.41

</h4>

0.40

↵

0.40

0.39

 again

0.37

↵↵↵

0.37

0.36

Activations Density 0.027%

patterns of emphasis or separation in the text formatting

No Comments

No Known Activations