INDEX

Explanations

references to reading or browsing content on the internet

oai_token-act-pair · gpt-4o-mini Triggered by @bot

Tokens preceding titles or descriptions

np_acts-logits-general · gemini-2.0-flash

describe features

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 oprot

-0.62

SaveChangesAsync

-0.51

SequentialGroup

-0.50

 >=",

-0.50

DoubleQuotes

-0.50

 StatelessWidget

-0.46

balleur

-0.46

 OnClickListener

-0.46

 nakalista

-0.46

 Branch

-0.45

POSITIVE LOGITS

 post

1.29

post

1.12

Post

1.05

<eos>

1.01

 Post

1.00

 пост

0.87

 pós

0.82

 posts

0.82

 Posts

0.78

Source

0.76

Activations Density 0.099%

references to reading or browsing content on the internet

Tokens preceding titles or descriptions

describe features

No Comments

No Known Activations