INDEX

Explanations

instances of the word "had" and variations of it in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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

had

-0.71

Had

-0.66

Had

-0.63

 Administrativna

-0.56

 actualidad

-0.55

had

-0.54

HAD

-0.51

WebElementEntity

-0.50

今は

-0.49

تقاوى

-0.48

POSITIVE LOGITS

 hoped

1.27

 originally

1.16

 previously

0.90

 intended

0.89

 initially

0.88

 ursprünglich

0.88

 planned

0.87

 earlier

0.86

originally

0.85

 originalmente

0.82

Activations Density 0.167%

instances of the word "had" and variations of it in different contexts

No Comments

No Known Activations