INDEX

Explanations

phrases that indicate a degree of measurement or extent in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 giovan

-0.65

 <>",

-0.65

 muzik

-0.65

Dum

-0.62

izop

-0.62

Dum

-0.59

vand

-0.58

 Lyman

-0.58

muz

-0.58

 Daar

-0.57

POSITIVE LOGITS

 degrees

1.29

 Degrees

1.26

 DEGREE

1.21

 Degree

1.17

 degree

1.17

GREES

1.17

Degree

1.15

Degrees

1.13

degrees

1.07

degree

1.06

Activations Density 0.123%

phrases that indicate a degree of measurement or extent in various contexts

No Comments

No Known Activations