INDEX

Explanations

phrases related to openness or the concept of being open in various contexts

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

 open

-0.77

 invokingState

-0.61

 ErrIntOverflow

-0.59

 ModelExpression

-0.59

 viewDidLoad

-0.57

open

-0.57

ΗΣ

-0.57

DoubleQuotes

-0.57

 HasFactory

-0.57

 CreateTagHelper

-0.56

POSITIVE LOGITS

invokeLater

0.75

 médicale

0.64

extrême

0.59

醐

0.55

SALT

0.55

tableFuture

0.55

océan

0.55

zess

0.55

 eccl

0.55

ensa

0.54

Activations Density 0.041%

phrases related to openness or the concept of being open in various contexts

No Comments

No Known Activations