INDEX

Explanations

actions related to assistance and improvement

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

followed by action verbs

np_acts-logits-general · gemini-2.0-flash

verbs of action

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.28.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

ectoria

-0.45

nativeElement

-0.43

 Autorizaciones

-0.39

esgue

-0.35

waitKey

-0.35

GORITH

-0.34

NamedQuery

-0.32

 okuyayım

-0.32

writeFieldEnd

-0.32

 autocollant

-0.32

POSITIVE LOGITS

AddTagHelper

0.87

 يتيمه

0.71

findpost

0.68

Билгалдахарш

0.65

Autoritní

0.62

Personendaten

0.60

:✨

0.59

 Numerade

0.57

 appunto

0.57

parsedMessage

0.56

Activations Density 0.501%

actions related to assistance and improvement

followed by action verbs

verbs of action

No Comments

No Known Activations