INDEX

Explanations

numerical patterns or sequences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

 चीज़ों

-0.92

 Italijani

-0.90

 Monfieur

-0.86

 &___

-0.84

setVerticalGroup

-0.82

 '\\;'

-0.82

 unknownFields

-0.79

 VIAF

-0.78

bootstrapcdn

-0.78

 fhew

-0.78

POSITIVE LOGITS

vit

0.48

0.42

que

0.42

сько

0.41

 volt

0.41

 lade

0.41

 им

0.40

Им

0.40

 Still

0.40

stretchr

0.39

Activations Density 0.213%

numerical patterns or sequences

No Comments

No Known Activations