INDEX

Explanations

frequency-related terms specifically associated with rates or averages

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

classnames

-0.55

 again

-0.50

Géographie

-0.49

外部リンク

-0.48

jahtera

-0.47

زيد

-0.47

Bodies

-0.45

setImageResource

-0.45

 weer

-0.44

addComponent

-0.44

POSITIVE LOGITS

per

0.98

per

0.95

 Ogni

0.92

Chaque

0.90

Ogni

0.89

PER

0.87

 mỗi

0.86

 प्रति

0.84

 bawat

0.83

every

0.82

Activations Density 0.239%

frequency-related terms specifically associated with rates or averages

No Comments

No Known Activations

frequency-related terms specifically associated with rates or averages

No Comments

No Known Activations