INDEX

Explanations

phrases that describe classifications or categories of things

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 cauſe

-0.69

 frequenza

-0.67

 purpoſe

-0.66

 iload

-0.64

 unut

-0.63

 foramen

-0.62

 pleaſure

-0.60

 exceptionnelle

-0.60

 Eternal

-0.60

 braccia

-0.60

POSITIVE LOGITS

Kinda

1.42

 sorta

1.36

 Kinda

1.32

kinda

1.23

 kinda

1.18

Somewhat

0.98

somewhat

0.95

 Somewhat

0.90

 ostavi

0.85

 somewhat

0.84

Activations Density 0.092%

phrases that describe classifications or categories of things

No Comments

No Known Activations