INDEX

Explanations

phrases indicating social interactions or opinions about people

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

Followed by punctuation or conjunctions

np_acts-logits-general · gemini-2.0-flash

popular, stuck, camera, slavery, static

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

RectangleBorder

-0.65

Appro

-0.52

Personendaten

-0.50

CloseOperation

-0.48

ScopeManager

-0.46

󠁮

-0.46

cantara

-0.45

rehension

-0.45

 fallu

-0.45

 Tense

-0.45

POSITIVE LOGITS

OGND

0.41

AutoScaleMode

0.37

 мәкал

0.35

 оригіналу

0.33

 تانيه

0.33

 utafitiHapana

0.33

cakup

0.32

 ويكيميديا

0.30

 layak

0.29

辜

0.29

Activations Density 0.330%

phrases indicating social interactions or opinions about people

Followed by punctuation or conjunctions

popular, stuck, camera, slavery, static

No Comments

No Known Activations