INDEX

Explanations

phrases that indicate conditions or contexts, often leading to conclusions or scenarios

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

AnchorTagHelper

-0.55

MessageOf

-0.55

opsida

-0.52

 estekak

-0.49

InjectAttribute

-0.45

SequentialGroup

-0.45

脚注の使い方

-0.44

TagMode

-0.43

PerformLayout

-0.43

⤹

-0.42

POSITIVE LOGITS

DrawerToggle

0.47

Aholisi

0.41

ので

0.40

Collectors

0.40

Since

0.40

zzleHttp

0.40

ViewChild

0.38

basicConfig

0.38

so

0.36

攘

0.36

Activations Density 0.608%

phrases that indicate conditions or contexts, often leading to conclusions or scenarios

No Comments

No Known Activations