INDEX

Explanations

numeric values and possibly identifiers in a formatted text

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

Bottom

-0.47

up

-0.45

AsUp

-0.45

存于互联网档案馆

-0.43

寡

-0.42

od

-0.41

 rowspan

-0.41

Faj

-0.41

inc

-0.40

 invest

-0.40

POSITIVE LOGITS

Personendaten

0.80

 propOrder

0.75

 nakalista

0.73

帖最后由

0.73

 <=",

0.72

findpost

0.69

 createState

0.69

UrlResolution

0.69

UnknownFieldSet

0.68

PerformLayout

0.66

Activations Density 0.950%

numeric values and possibly identifiers in a formatted text

No Comments

No Known Activations