INDEX

Explanations

various forms of the phrase "take action."

oai_token-act-pair · gemini-2.0-flash

take

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

take

-1.25

Take

-1.23

 take

-1.21

 Take

-1.17

 taken

-1.16

taken

-1.13

taking

-1.05

 TAKE

-1.02

 taking

-1.02

Taking

-1.00

POSITIVE LOGITS

 resourceCulture

0.69

rungsseite

0.59

berspace

0.50

ÉN

0.50

 steps

0.49

 Back

0.47

 charge

0.47

CodedInputStream

0.47

úcar

0.47

 back

0.47

Activations Density 0.821%

various forms of the phrase "take action."

take

No Comments

No Known Activations