INDEX

Explanations

contrasts and qualifiers in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 35-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_35/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.35.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

IntoConstraints

-0.46

 tidak

-0.45

 modelAndView

-0.45

 Roskov

-0.45

новниш

-0.44

GEBURTSDATUM

-0.44

 nemus

-0.43

 nicht

-0.41

 δεν

-0.40

not

-0.40

POSITIVE LOGITS

 rather

2.25

Rather

1.98

 Rather

1.92

rather

1.87

 vielmehr

1.86

 plutôt

1.79

 piuttosto

1.76

むしろ

1.67

 скорее

1.60

 raczej

1.48

Activations Density 0.410%

contrasts and qualifiers in the text

No Comments

No Known Activations