INDEX

Explanations

expressions of agreement or strong opinions

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

following pronouns or question words

np_acts-logits-general · gemini-2.0-flash

program name and description

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

 AssemblyCulture

-0.81

AndEndTag

-0.78

SourceChecksum

-0.78

</caption>

-0.76

 الرياضيه

-0.74

!")

-0.72

momix

-0.70

SOUNDBITE

-0.68

^(@)

-0.68

 ddelweddau

-0.67

POSITIVE LOGITS

 What

0.77

Honestly

0.76

 Honestly

0.75

Why

0.73

What

0.72

 Wasn

0.70

0.68

Looks

0.68

 Maybe

0.67

Activations Density 0.155%

expressions of agreement or strong opinions

following pronouns or question words

program name and description

No Comments

No Known Activations