INDEX

Explanations

technical or specialized terms related to biological or computational fields

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

followed by the letter "s"

np_acts-logits-general · gemini-2.0-flash

homologs and beliefs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ostante

-0.87

脚注の使い方

-0.81

 '\\;'

-0.80

 defaultstate

-0.78

childs

-0.77

 مشين

-0.76

 Chwiliwch

-0.75

Manbalar

-0.75

RetentionPolicy

-0.75

dersfield

-0.75

POSITIVE LOGITS

2.33

ی

1.56

1.43

er

1.36

es

1.33

ه

1.30

1.27

1.23

的

1.21

1.19

Activations Density 2.026%

technical or specialized terms related to biological or computational fields

followed by the letter "s"

homologs and beliefs

No Comments

No Known Activations