INDEX

Explanations

specific features or identifiers related to a particular system or software, likely in a technical or programming context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_65k/average_l0_73

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.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

-0.52

arbox

-0.48

ormal

-0.47

 Tampa

-0.47

 dead

-0.47

 Landwir

-0.46

 Lubbock

-0.46

gå

-0.46

@}

-0.45

 living

-0.45

POSITIVE LOGITS

ES

1.13

ES

0.91

endphp

0.83

oa̍t

0.77

 calendriers

0.75

ESI

0.73

]';

0.72

esk

0.71

es

0.71

évaluateur

0.71

Activations Density 0.045%

specific features or identifiers related to a particular system or software, likely in a technical or programming context

No Comments

No Known Activations