INDEX

Explanations

references to specific articles or source materials

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_11/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

OGND

-0.69

 صوتيه

-0.68

 Италијани

-0.64

MessageOf

-0.64

parsedMessage

-0.64

 IBOutlet

-0.59

المناصب

-0.58

 doInBackground

-0.58

thâu

-0.58

GEBURTS

-0.57

POSITIVE LOGITS

the

0.59

 usually

0.52

 most

0.47

 only

0.43

 often

0.43

 relatively

0.42

 actual

0.42

 sını

0.41

The

0.40

any

0.40

Activations Density 0.933%

references to specific articles or source materials

No Known Activations