INDEX

Explanations

words related to places and specific names, indicating geographical or historical relevance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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

writerow

-0.50

ja

-0.38

ods

-0.38

exp

-0.38

atts

-0.37

こし

-0.36

cub

-0.36

makeText

-0.36

pe

-0.36

AQ

-0.36

POSITIVE LOGITS

 zoude

0.63

libatkan

0.57

 dezelve

0.56

apatalk

0.55

visuel

0.54

 zelve

0.54

tiéndose

0.53

ésult

0.53

berdayakan

0.53

 ſch

0.52

Activations Density 0.332%

words related to places and specific names, indicating geographical or historical relevance

No Comments

No Known Activations