INDEX

Explanations

terms and phrases that convey authenticity or sincerity

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-mlp/layer_6/width_131k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.6.hook_mlp_out

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

 genuinely

-0.94

genu

-0.59

 legitimately

-0.55

 sincerely

-0.53

 seriously

-0.43

 Genu

-0.41

 cons

-0.39

 mechanically

-0.38

 actively

-0.38

 conveniently

-0.37

POSITIVE LOGITS

SourceChecksum

0.65

 AssemblyTitle

0.62

Selfer

0.56

rrggbb

0.56

يميديا

0.54

########.

0.54

 PeEnEo

0.54

WithIOException

0.54

новниш

0.53

RegressionTest

0.53

Activations Density 0.001%

terms and phrases that convey authenticity or sincerity

No Comments

No Known Activations