INDEX

Explanations

repeated instances of the token "num" associated with numerical values

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

Morfologia

-0.64

 financière

-0.59

 Angleterre

-0.59

étricas

-0.55

 Ελλά

-0.55

membered

-0.54

անի

-0.54

 mene

-0.53

 وتسجيلات

-0.53

'">

-0.52

POSITIVE LOGITS

num

3.39

num

3.19

Num

2.98

NUM

2.72

Num

2.69

NUM

2.38

getNum

1.92

setNum

1.90

 getNum

1.85

nums

1.83

Activations Density 0.068%

repeated instances of the token "num" associated with numerical values

No Comments

No Known Activations