INDEX

Explanations

proper nouns and specific identifiers related to people and places

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

fjspx

-0.54

ImageContext

-0.52

culous

-0.52

yelitis

-0.52

richTextPanel

-0.51

BagLayout

-0.51

ajur

-0.50

."],

-0.48

igurumi

-0.47

 Minaj

-0.47

POSITIVE LOGITS

enterian

0.52

 Италијани

0.50

которые

0.50

äsident

0.48

örgy

0.46

withstanding

0.45

apeake

0.42

nezeu

0.42

кобрита

0.42

persky

0.41

Activations Density 2.021%

proper nouns and specific identifiers related to people and places

No Comments

No Known Activations