INDEX

Explanations

code-related tokens such as assignment operators, variable names, `new`, and code block delimiters

oai_token-act-pair · gemini-2.0-flash

code/files

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_8/width_16k/average_l0_52

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.8.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

featureID

-0.63

Kaieteur

-0.59

mení

-0.52

endphp

-0.50

icestershire

-0.50

 Bär

-0.50

Tembelea

-0.50

 &___

-0.49

MigrationBuilder

-0.49

--}}

-0.49

POSITIVE LOGITS

WriteTagHelper

0.59

Демографія

0.58

ranean

0.54

Poloha

0.54

CompilerServices

0.54

 Occupations

0.53

 OBITUARY

0.53

واقع

0.52

Viki

0.51

Publication

0.51

Activations Density 5.297%

code-related tokens such as assignment operators, variable names, `new`, and code block delimiters

code/files

No Comments

No Known Activations