INDEX

Explanations

specific formatted text structures or organized data segments

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

"file" or data storage references

np_acts-logits-general · gemini-2.0-flash

specific lists or files

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

verifyException

-0.59

 miniaturka

-0.58

 wikipagina

-0.57

 يتيمه

-0.57

étoit

-0.54

 Biôgrafia

-0.49

harapkan

-0.48

ambién

-0.47

 Италијани

-0.47

 ilustracja

-0.46

POSITIVE LOGITS



0.46

Sax

0.41

sax

0.41

tanooga

0.40

mbic

0.39

MOM

0.39

Duck

0.39

Garden

0.38

Dad

0.38

Sax

0.38

Activations Density 0.137%

specific formatted text structures or organized data segments

"file" or data storage references

specific lists or files

No Comments

No Known Activations