INDEX

Explanations

terms related to legal and compliance contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_18/width_65k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.18.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 desiring

-0.69

 craving

-0.67

 defire

-0.66

 desire

-0.66

writeField

-0.65

endforeach

-0.65

 desires

-0.65

Referensi

-0.60

 yearning

-0.58

 thirst

-0.58

POSITIVE LOGITS

osoever

0.55

tské

0.52

badlogic

0.50

енча

0.50

=[[

0.46

getKeyCode

0.46

 Infórmanos

0.45

buffer

0.45

Spa

0.44

 block

0.43

Activations Density 0.094%

terms related to legal and compliance contexts

No Comments

No Known Activations