INDEX

Explanations

mentions of images and multimedia content

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 27-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_27/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.27.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ftagPool

-0.63

 pleaſure

-0.60

Jeografia

-0.59

lccn

-0.56

quelize

-0.56

 متحده

-0.54

 AssemblyVersion

-0.52

 propOrder

-0.51

Personensuche

-0.51

LElement

-0.51

POSITIVE LOGITS

AutoModerator

0.35

 BaseActivity

0.33

lan

0.33

afterEach

0.29

作

0.29

 getItemId

0.28

su

0.28

zwi

0.28

para

0.28

новниш

0.28

Activations Density 0.168%

mentions of images and multimedia content

No Known Activations