INDEX

Explanations

conditional and hypothetical phrases

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

CppMethod

-0.80

 Normdatei

-0.79

 uppermost

-0.68

Hentet

-0.68

 {}));

-0.66

(")");

-0.65

**/

-0.63

 للمعارف

-0.63

octanol

-0.62

nador

-0.62

POSITIVE LOGITS

 Suppose

0.75

Suppose

0.73

 suppose

0.69

0.60

 instead

0.58

?!

0.58

Imagine

0.56

 when

0.56

 that

0.55

 imagine

0.55

Activations Density 0.110%

conditional and hypothetical phrases

No Comments

No Known Activations