INDEX

Explanations

conjunctions and phrases that indicate a sequence or relationship between ideas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

Condol

-0.39

 {}));

-0.38

})()

-0.36

 samoz

-0.35

 imminent

-0.35

 rumput

-0.34

******/

-0.34

 puesta

-0.34

 InputDecoration

-0.34

 Yamamoto

-0.33

POSITIVE LOGITS

 ſte

0.74

 ſta

0.72

 myſelf

0.68

 poffe

0.67

 himſelf

0.65

 Majefty

0.63

 houſe

0.60

 ſhe

0.60

 perſon

0.59

 Monfieur

0.59

Activations Density 0.569%

conjunctions and phrases that indicate a sequence or relationship between ideas

No Comments

No Known Activations