INDEX

Explanations

actions involving filling, checking, or figuring out processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 33-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.33.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

Personendaten

-0.56

Ӕ

-0.54

GEBURTSDATUM

-0.54

رایی

-0.50

localctx

-0.50

AsUp

-0.49

SharedDtor

-0.49

Handlung

-0.48

 חיצוניים

-0.48

AnchorTagHelper

-0.47

POSITIVE LOGITS

out

0.51

 &___

0.49

off

0.46

 down

0.45

up

0.42

httphttps

0.40

 away

0.39

 campamento

0.37

AndEndTag

0.36

 vastaan

0.35

Activations Density 0.288%

actions involving filling, checking, or figuring out processes

No Comments

No Known Activations