INDEX

Explanations

tokens or symbols that indicate numerical or structured data representations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_7/width_16k/average_l0_92

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 Италијани

-0.88

UserScript

-0.77

transQ

-0.75

 OMITBAD

-0.73

клопе

-0.73

تقاوى

-0.72

 فريبيس

-0.71

stateMutability

-0.70

exitRule

-0.68

SharedDtor

-0.68

POSITIVE LOGITS

 bailarina

0.33

válasz

0.31

sipasi

0.30

fören

0.30

erçe

0.29

gjenge

0.29

 TÉCN

0.29

lateinit

0.29

Bakgrunnsstoff

0.29

delning

0.28

Activations Density 1.143%

tokens or symbols that indicate numerical or structured data representations

No Comments

No Known Activations