INDEX

Explanations

highly favorable or prestigious attributes related to various subjects

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

Preceding words of prepositional phrases

np_acts-logits-general · gemini-2.0-flash

greatest of all time

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Personendaten

-0.65

+#+#

-0.55

adpleegd

-0.47

üyada

-0.46

:✨

-0.46

IsMutable

-0.46

\{\\

-0.46

ContentLoaded

-0.46

PerformLayout

-0.45

 wichtiger

-0.43

POSITIVE LOGITS

 ever

1.39

 imaginable

1.09

 EVER

1.08

ever

1.00

 ooit

0.89

 jemals

0.89

Ever

0.87

 Ever

0.87

EVER

0.84

 überhaupt

0.73

Activations Density 0.172%

highly favorable or prestigious attributes related to various subjects

Preceding words of prepositional phrases

greatest of all time

No Comments

No Known Activations