INDEX

Explanations

references to online moderation and conduct guidelines

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 17-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_17/width_65k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.17.hook_resid_post

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

PositiveButton

-0.61

getJSONArray

-0.59

 विश्वसनीयता

-0.58

 DialogInterface

-0.56

 >=",

-0.54

urnia

-0.53

клопе

-0.52

 aDecoder

-0.52

 kaynağından

-0.52

NewUrlParser

-0.52

POSITIVE LOGITS

 biografias

0.48

 Administrator

0.48

featureID

0.48

 hát

0.48

Personensuche

0.47

ⓧ

0.45

Any

0.45

zoo

0.44

Dune

0.44

Enllaces

0.44

Activations Density 0.044%

references to online moderation and conduct guidelines

No Comments

No Known Activations