INDEX

Explanations

instances of code import statements, particularly in Java and Android

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

 Paglinawan

-0.93

Vidite

-0.85

InputBorder

-0.84

GOTREF

-0.82

Viited

-0.79

Jeografia

-0.79

RectangleBorder

-0.78

UIControlState

-0.77

pulumi

-0.76

tanleria

-0.74

POSITIVE LOGITS

DialogResult

0.42

<u>

0.38

 dayan

0.37

 παρά

0.37

 Young

0.36

 chromatin

0.36

ovi

0.36

 ordinaires

0.35

ritage

0.35

accum

0.35

Activations Density 0.009%

instances of code import statements, particularly in Java and Android

No Comments

No Known Activations