INDEX

Explanations

phrases that indicate knowledge or awareness of specific information or subjects

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

Bene

-0.64

 Bene

-0.59

Each

-0.57

 assemb

-0.56

 hypocrite

-0.56

Yap

-0.55

 bình

-0.54

 diri

-0.54

 presided

-0.53

卑

-0.53

POSITIVE LOGITS

 للمعارف

0.95

 adanya

0.90

Tikang

0.82

elemField

0.78

 gibts

0.76

an

0.72

DotNetBar

0.71

 تانيه

0.71

 مرئيه

0.71

Activations Density 0.295%

phrases that indicate knowledge or awareness of specific information or subjects

No Comments

No Known Activations