INDEX

Explanations

technical terms and measurements related to physics and engineering

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Notably

-0.55

ittäin

-0.52

 kısm

-0.52

 cadeira

-0.52

 vegetales

-0.50

 būtų

-0.49

 autobuses

-0.49

leyeb

-0.48

روض

-0.48

曖昧さ回避

-0.48

POSITIVE LOGITS

Personensuche

0.93

 виправивши

0.91

IntoConstraints

0.88

 AssemblyCulture

0.87

awtextra

0.86

posedge

0.77

AddTagHelper

0.76

ViewFeatures

0.73

AccessorTable

0.68

cshtml

0.67

Activations Density 19.204%

technical terms and measurements related to physics and engineering

No Comments

No Known Activations