INDEX

Explanations

phrases related to moral dilemmas or ethical considerations

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

consequences and conditions

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 surla

-0.50

writeFieldEnd

-0.46

 continúas

-0.45

 سرا

-0.44

 للمعارف

-0.43

 autorytatywna

-0.43

ⓧ

-0.43

Kariera

-0.43

 wireType

-0.43

 begitu

-0.43

POSITIVE LOGITS

SBATCH

0.48

CompilerServices

0.42

LookAnd

0.35

 magari

0.35

SwingConstants

0.35

 FetchType

0.35

LEGGI

0.34

GEBURTSDATUM

0.34

jalá

0.33

WindowConstants

0.33

Activations Density 0.025%

phrases related to moral dilemmas or ethical considerations

consequences and conditions

No Comments

No Known Activations