INDEX

Explanations

occurrences of the word "по."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_22/width_16k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

########.

-0.73

caloosa

-0.57

ไหน

-0.54

 شيئ

-0.54

 volna

-0.53

 Cæsar

-0.52

 őket

-0.52

dotta

-0.51

findung

-0.51

__':

-0.51

POSITIVE LOGITS

ThroughAttribute

0.64

EndInit

0.56

webElementXpaths

0.54

 Solidity

0.54

bitField

0.54

 CreateTagHelper

0.54

MigrationBuilder

0.53

MediaStore

0.53

GraphicsUnit

0.52

WithIOException

0.52

Activations Density 0.019%

occurrences of the word "по."

No Comments

No Known Activations