INDEX

Explanations

words starting with "bla" or "blu".

oai_token-act-pair · gemini-2.0-flash

repeated patterns in text

np_token-act-pair-logits · gpt-4o-mini

bla

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

Ljava

-0.44

diam

-0.43

 investi

-0.43

尊

-0.41

 Wahl

-0.40

yr

-0.40

rund

-0.40

ventil

-0.40

pyx

-0.39

 kiện

-0.39

POSITIVE LOGITS

enderror

0.90

帖最后由

0.80

IsContent

0.80

Архівовано

0.79

 propOrder

0.78

ContentAsync

0.76

ArgsConstructor

0.74

contentLoaded

0.73

IndentedString

0.72

 AssemblyProduct

0.70

Activations Density 0.247%

words starting with "bla" or "blu".

repeated patterns in text

bla

No Comments

No Known Activations