INDEX

Explanations

a wide range of words that do not show a strong consistent theme

oai_token-act-pair · gemini-2.0-flash

French language text

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_4/width_16k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

Efq

-1.30

 itſelf

-1.16

 houſe

-1.16

 myſelf

-1.03

 للمعارف

-1.00

StoryboardSegue

-0.99

 Houſe

-0.97

 juſ

-0.97

 ſte

-0.96

 pleaſure

-0.96

POSITIVE LOGITS

0.45

<eos>

0.45

cu

0.44

us

0.43

amer

0.42

</strong>

0.41

 River

0.41

top

0.40

}}{\

0.40

Activations Density 4.772%

a wide range of words that do not show a strong consistent theme

French language text

No Known Activations