INDEX

Explanations

numeric values or identifiers related to a database or coding context

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

 jäsen

-0.53

 Dabei

-0.53

 zijne

-0.52

 aikana

-0.52

 esimer

-0.52

Dabei

-0.50

 aihe

-0.49

 enggak

-0.49

 beschik

-0.48

 navideña

-0.47

POSITIVE LOGITS

to

0.63

########.

0.60

To

0.53

 kaarangay

0.53

To

0.50

IndentedString

0.49

numerusform

0.48

 AssemblyCulture

0.48

 ujednoznacz

0.48

 utafitiHapana

0.48

Activations Density 0.066%

numeric values or identifiers related to a database or coding context

No Comments

No Known Activations