INDEX

Explanations

words and phrases related to measurements and sizes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 25-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_25/width_16k/average_l0_116

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.25.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

 heaviest

-0.86

 Larger

-0.82

 Faster

-0.82

Larger

-0.82

Faster

-0.82

largest

-0.81

faster

-0.80

Higher

-0.80



-0.80

 fastest

-0.79

POSITIVE LOGITS

 really

0.84

 very

0.82

 extremely

0.81

 super

0.81

 incredibly

0.79

 insanely

0.75

 unbelievably

0.73

 REALLY

0.71

 ridiculously

0.69

 ultra

0.68

Activations Density 0.435%

words and phrases related to measurements and sizes

No Comments

No Known Activations