INDEX

Explanations

references to films and their associated awards

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-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.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

InvalidProtocol

-0.65

 newState

-0.57

ỡng

-0.52

Géographie

-0.51

География

-0.49

 szcze

-0.49

 femmin

-0.47

存于互联网档案馆

-0.47

جغرافيا

-0.46

ILLUSTRATION

-0.46

POSITIVE LOGITS

pieler

0.71

ⓘ

0.70

 препратки

0.67

 تضيفلها

0.65

 ujednoznacz

0.63

клопе

0.61

tslib

0.59

tissimi

0.56

了解更多

0.56

 famí

0.56

Activations Density 0.246%

references to films and their associated awards

No Comments

No Known Activations