INDEX

Explanations

references to conferences or academic proceedings in the field of computer science or related topics

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



-0.62

 }}"></

-0.59

SharedCtor

-0.56

 AppModule

-0.55

NewGuid

-0.55

 صوتيه

-0.54

 дописавши

-0.52

 resourceCulture

-0.52

 andererseits

-0.52

 crystallization

-0.52

POSITIVE LOGITS

GetAxis

0.51

Бахар

0.51

thest

0.50

Axis

0.49

dic

0.47

ագրություններ

0.46

 django

0.46

省市镇

0.46

 Illuminate

0.45

chyma

0.45

Activations Density 0.064%

references to conferences or academic proceedings in the field of computer science or related topics

No Comments

No Known Activations