INDEX

Explanations

phrases related to error messages regarding image insertion or manipulation

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

Beginning of non-English sentences

np_acts-logits-general · gemini-2.0-flash

consuming or calculated

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

MigrationBuilder

-0.79

 pleaſure

-0.72

 CreateTagHelper

-0.71

 itſelf

-0.70

momix

-0.70

 greateſt

-0.69

ⓧ

-0.69

NameInMap

-0.67

存于互联网档案馆

-0.66

 ſind

-0.66

POSITIVE LOGITS

sa

0.45

at

0.40

ste

0.38

be

0.37

man

0.35

Navigation

0.35

pat

0.35

is

0.34

nor

0.34

attribution

0.33

Activations Density 0.099%

phrases related to error messages regarding image insertion or manipulation

Beginning of non-English sentences

consuming or calculated

No Comments

No Known Activations