INDEX

Explanations

structured data representations, including tuples and lists with specific numeric values

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

 Савезне

-0.56

 Numerade

-0.47

 Мексичка

-0.46

ERVIS

-0.46

囗

-0.45

Билгалдахарш

-0.42

تقاوى

-0.41

 jLabel

-0.40

subtotal

-0.40

 '\\;'

-0.39

POSITIVE LOGITS

0.51

--

0.46

featureID

0.46

–

0.44

$-\

0.42

-.

0.41

$-

0.40

–

0.40

--

0.40

 propOrder

0.40

Activations Density 1.682%

structured data representations, including tuples and lists with specific numeric values

No Comments

No Known Activations