INDEX

Explanations

This neuron seems to activate on a grab bag of different things, and it's not clear what the common thread is. Some of them seem related to CSS code, and another group relates to emotions. I'm going to guess it is looking for emotional or demonstrative language.

oai_token-act-pair · gemini-2.0-flash

Indicating/Telling

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_19/width_16k/average_l0_12

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.19.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

-0.43

dut

-0.38

 pelo

-0.38

Captor

-0.38

 manca

-0.38

 Screenshots

-0.37

 pous

-0.37

 Maus

-0.36

sights

-0.36

farwyddwr

-0.36

POSITIVE LOGITS

 telling

1.40

 indicating

1.23

 tells

1.23

 saying

1.22

 tell

1.21

 indicate

1.20

tells

1.15

indicate

1.14

 indicates

1.14

 INDIC

1.11

Activations Density 2.364%

This neuron seems to activate on a grab bag of different things, and it's not clear what the common thread is. Some of them seem related to CSS code, and another group relates to emotions. I'm going to guess it is looking for emotional or demonstrative language.

Indicating/Telling

No Comments

No Known Activations

This neuron seems to activate on a grab bag of different things, and it's not clear what the common thread is. Some of them seem related to CSS code, and another group relates to emotions. I'm going to guess it is looking for emotional or demonstrative language.

Indicating/Telling

No Comments

No Known Activations