INDEX

Explanations

numeric values, including references, counts, and measurements

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

tr

-0.40

备注

-0.39

 Administrativna

-0.37

 renov

-0.36

 Renov

-0.35

 jLabel

-0.35

novu

-0.35

LikeLike

-0.35

 проєкту

-0.35

region

-0.35

POSITIVE LOGITS

 defaultstate

0.71

+#+

0.54

WritableDatabase

0.53

OGND

0.48

IsMutable

0.46



0.46

endcsname

0.45

 bezeichneter

0.45

 ویکی‌پدی

0.44

دانشنامهٔ

0.43

Activations Density 0.079%

numeric values, including references, counts, and measurements

No Comments

No Known Activations